In this course, we will review several classes of robots ranging from older more traditional robots

to newer types and even future types of robots. There's a strong interest in enabling both older
and newer robots to work safely with people. Early robots all worked in factories performing
heavy and dangerous jobs, such as welding, palletizing. In the 1980s, smaller robots emerged for
use in assembly and packaging applications. Around the year 2000, we started to see a few robots
developed for non-factory applications, such as surgery and service applications. By 2010, the real
market emerged for autonomous mobile vehicles, some of which carried robots. Also around this
time, robots were being developed for either autonomous or master slave work in hazardous
environments. Although still in the research phase, we're now seeing some humanoid robots that
can walk around and grasp items. Automation has evolved dramatically over the years. I mean, the
first robots were hydraulic, very limited capabilities, very small programs, we're talking in the late
'50s early '60s first robotic system. It took 10 years to even get beyond that point. Then, all of a
sudden through the '70s robotic started growing and the arm shapes are changing and more kinds
of industrial robots started evolving. By the '80s, industrial robotics were going strong, but the
development, the technology by about the mid '80s slowed down. The controllers got faster, the
wiring, and controls made him a little bit more safe in terms of the number circuits and additional
functionalities for applications are added. Industrial robotics focused on application-based; paint
system, welding systems, palletizing systems. So, the robots are getting cheaper, faster, and more
capable to handle more applications. But the overall functionality remained the same. You bought
a robot for an application, you put it in a cage, you tied your safeties into it, and you'd let it do the
work. It stayed like this all the way up through about 2008. About 2008, there's a company in
Denmark, actually two companies at this time, were developing technologies to change the way
robotics would be done. In 2008, the Universal Robots sold their first five-kilogram arm to a
company. The arm didn't need fencing. It didn't need a lot of the things that all the current robotic
systems did need industrial robots. It was safe to work around people, have special software and
feedback to allow it to work safely. If it did collide with somebody, it did it in a safe manner. It was
controlled. This was revolutionary.

Play video starting at 3 minutes 5 seconds and follow transcript3:05

Now, that was over in Denmark. In the US and other company, Rethink Robotics was developing
similar ideas, but instead of taking a software and electrical controls point of safety, they did theirs
by more mechanical and elasticity in the joints. But they focus more heavily on a user interface
making it friendly, where Universal focus there is on streamlining their system. By 2012, both
these companies were selling product in the US. Universal introduced a 10 kilogram arm, and
things changed. We're in 2018, six years later, up to that point, we had about maybe 14 companies
industrial robotics. Now, we have those same 14 companies, but now there's another 20 new
robotic companies that have come out using this new technology, and they're springing up
constantly in the price which was holding steady for years and years and years, started to
plummet. It started with universal and as these more of these new companies start coming into
the marketplace, it's dropping further, faster. With the cade systems, the robots in the early 80s, it
was a single channel of safety. So, if you open up a gate, it would open up a circuit in the robot and
would shut down the motors and kill power. By mid '90s, it became mandatory that you had to
have tools to two circuits. So, you had wires coming back from a switch on a fence, and when they
open the fence, they would open the circuit, it would kill power to the robot. There was also teach
mode, you could have a gate circuits to allow the robot to still move but somebody had to have
their hand on a pendant. So they had to have safety in their hands. So if the system wasn't going
automatic with their teach pendant, they had to have safety in their hand. The newer systems
have changed that way of operation. All the robots up until about 2008, you had to have to a teach
pendant. The reason is there was enabling switches on the back of that pendant or on the side of
the pendant, or somewhere in the pendant where they had to have a grip on constantly. If they
squeeze too hard or they let go, the robot would shut down for teaching a robot. With the new
systems that came out, Universal started it simply. They went down to a single button, but with
their software, they're able to hand guide the robot and not have to use a pedant to move around,
and they didn't need their hands a little switches. As robots have been evolving up until now, now
you see robotic systems going completely away from teach pendants, where you may have a
tablet or a laptop that you're using to help program; or using hand guidance, whereas before will
be unheard of to think about grabbing a big robotic arm and move it around. But these days, it's
becoming more and more common and are getting into more processes where this hand guidance
is capable. Now, they had hand guidance back needs too but it's a old crude system using steering
arm set controls. But this has become more elegant, more easier, and it's getting to the point
where it almost takes very little effort with the feedback systems to move their arms around
position where you need to be.

Let's take a look at a variety of industrial robots. Here we see two small six-axis robots. The larger
one has a payload of three kilograms, and the smaller one has a payload of 500 grams. Next is a
range of four-axis robots designed to pick up a part, rotate it and place it. This class of robot can
often move very quickly with a pick and place cycle of 300 to 500 milliseconds, with payloads of a
few kilograms. Tool speeds can reach 10 meters per second, making these small robots very
dangerous. Another example of an industrial robot is a large six-axis robot, spot welding line.
These robots typically have 100 kilogram payload, so obviously this is a very powerful and
dangerous robot. A six-axis collaborative robot can also work together with the person who is
loading a part onto the robot end-effector, prior to the robot installing the part in a difficult to
access location within the automobile. Here's a line of two armed robots working in a factory in
Japan that makes change machines. Recently, there has been some interest in humanoid robots
that can work with two arms to reduce the need for fixtures to hold parts. Now, let's take a look at
some examples of service robots. Here, we have the Da Vinci surgical robot from Intuitive Surgical.
This robot is a master slave robot where the surgeon is sitting at the table and has a 3D virtual
reality headset that displays a stereo image from two cameras, that are mounted on the ends of
laparoscopic tubes inserted inside the patient's body. The surgeon manipulates two forceps styles
grippers which control the motion of miniature tools on the ends of the two more laparoscopic
tubes inside the body. Hundreds of these systems are now in use worldwide performing a variety
of operations. Here we see something completely different, a large industrial robot that has been
converted to an amusement park ride, waving around in the air, some either very brave or very
foolish people. Next, we see a hospital service robot that has enough strength to carry a patient
around. This is only a concept at this time. The last service robot we will present is a fruit picking
robot working in an orchard. There's a fair bit of interests and developing robots that can work
outdoors, in dirt and mud, and perform various tedious agricultural tasks such as picking, weeding,
transplanting and spraying. Robots are now rolling around and in a few cases walking around.
Precise automation sells a collaborative robot to Adept Technology, who mounts it on their
Autonomous Mobile Base which carries around a box containing semiconductor wafers from tool
to tool in semiconductor factories. This is an example of a robot that can service multiple tools 24
hours a day. Here you see the Kiva, autonomous vehicle, delivering picking racks to a human picker
in an Amazon warehouse. An elaborate planning system tells a bunch of mobile robots which racks
to bring to a human picker. So, an order can be picked at one location without the need for the
person to run all over the warehouse. This dramatically improves the efficiency of the pickers.
Serving robots are now emerging at some fast food restaurants in Japan. Customers first place
their orders on tablets at their table and then the serving robot rolls up with a tray, which the
customer removes and returns when done. This mobile base has omnidirectional wheels. If these
wheels with rollers all turn in the same direction, the robot moves forward. If they turn in opposite
directions, the mobile base moves sideways, and can therefore move in any vector without
steering. The robot is able to load items onto the platform or place them into a bin. Robots are
increasingly being used to work in very dangerous environments that humans cannot enter. Most
people are familiar with the Mars rover which wanders around and takes soil samples. Here's
some less well-known types of robots that enter hazardous environments. This is an underwater
seafloor mining robot, developed to mind manganese nodules and other minerals from the
seafloor. Note the size of the person standing next to this huge machine. Here's a firefighting
robot which can move into much hotter and more dangerous situations than a human firefighter.
This is an autonomous underwater survey robot which can follow a planned route and collect data
for some weeks at a time without control from the surface, and here is a very scary robot, a
military robot that is remotely controlled by a person. However, some planners are suggesting
military robots might make decisions by themselves which raises a lot of very serious ethical
questions. Finally, when we think about robots interacting with humans, one of the most
compelling forms of robot is the Android or humanoid type robot. Mentally, we equate a human
shape with human abilities including perception, reasoning, communication and manipulation.
Android-type robots had been a topic of fascination in Japan for many years, although less so in
the US and Europe. Honda has had a program which has received a fair bit of publicity with an
Android they call Asimo. You can see the evolution of this design moving from a pair of legs used
to study walking and balance to a complete robot with some simple multi-finger hands. This robot
really can't do anything useful yet other than walk around. After about 20 minutes, it's battery
runs out and it falls over. In summary, there are many classes of robots emerging, and most of the
newer types of robots will be interacting with people. So, understanding collaborative robots is
very important.

Robots are continually evolving and most of the newer type of robots are interacting with people.
As such, there has been an enormous demand in the industry to enhance the safety of
personnel working with these collaborative robots. The objective of safety is to prevent
the occurrence of an injury, harm or danger. Safety is essential all the way from the
enterprise to the operator level. The absence of safety can cause damage, death, financial
loss, reputation and productivity. However, with proper planning, training and execution,
safety can be practiced and maintained at all times. As you'll see throughout the course, it
takes a proactive safety approach. We build upon a 3E or safety pyramid which lays safety
fundamentals, meters safety performance and continually provides the opportunities for
improvement. The safety pyramid has the 3E tools. The most effective, elimination, refers
to physically removing a hazard. Engineering controls which incorporate programming,
safety curtains, et cetera, to prevent danger. The third here is enterprise controls which
changes the way an operator walks through walk instructions, restrictions, et cetera.
Several continually emerging robot safety standards provide the framework to enhance
safety. ISO 10218-1, 2011 Part One talks about the safety essential for industrial robots. It
addresses robot manufacturers. The standard evolved from the 2006 edition and includes
the concept of risk assessment at the application level. ISO 10218-1, 2011 Part Two
incorporates the safety requirements at the integrator level. The ISO/TS 15066:2016 builds
upon the two mentioned standards and provides safety requirements for collaborative
industrial robots. Although this technical standard primarily targets collaborative industrial
robots, there's a cross transfer of knowledge applicable to traditional industrial robots.
The Robotics Industries Association, RIA, put together a technical report in 2016, the
RIA/TR R15.606-2016. The report enhances the safety of personnel with collaborative
industrial robot systems and represents the primary focus of the course. Although a
technical report, it's continually improving and is expected to mature into a standard by
early 2020. Moving on to control reliability of the robots. Controlled reliability is a concept
and represents a function of relative safety. However, the reliability is difficult to quantify.
A matrix provides a method to validate a robot or robotic system design. The performance
level varies from b to e, and structure categories from one to four. Both represent from
negligible to high risks which we will get into more detail later. The robots are expected to
meet the safety requirements of performance level d, with the structure of category 3.
Based on the input of a risk assessment, it may require higher control reliability based on
the probability of occurrence of a hazard. At this point, it will be helpful to understand
what distinguishes a collaborative robot from a non-collaborative robot. Traditionally, an
industrial robot would exclude operator axis to the operations area while the robot is still
active. A robot is usually protected with a fence around it to prevent human interference.
Now, in an instance where an operator does enter the cell without hitting stop button or
absence of present detector, the robot can cause harm to the operator since this does not
come to a protective stop. On the other hand, collaborative robots have inherent safety
features that can detect a threshold contact force and can instantly stop. In these
collaborative workspaces, robots exhibit precision and power and humans can solve
imprecise problems. Such a collaborative application finally affords robots and humans to
share the same workspace. We introduced the critical difference between a collaborative
and non-collaborative operation. Let's dig deeper into different kinds of operation based
on a function of time and distance between the operator and a robot. This foundation of
different kinds of operations will be beneficial as we progress through different facets of
the course. This understanding will help you recognize and design the type of operations
 that you would like to incorporate in your application. The three types of robot operations
are coexistence, cooperation and collaboration. In a coexistence type of operation, a robot
and an operator function at different times and in separate workspaces. A protective fence
surrounds the robot and access doors are locked. This fence prevents human access while
the robot operation is active. However, an operator may stop the robot operation and
enter for required maintenance in industrial robots. In a cooperative type of operation, an
operator and a robot work in the same workspace. However, the different stages of the
tasks are realized at various times during the operations. These operations require
adequate safety measures in place, for example, progressive padding, pressure mats, et
cetera. These measures prevent operator access to the restricted area of robot operations
as it may cause potential harm or an injury. Loading and unloading off parts is an example
of a cooperative operation. Now, on to our final type of operation, the collaborative
operation. In this type, an operator and a robot perform tasks within the same workspace.
Unlike cooperative, a collaborative operation is realized at the same work time. The
collaborative activity is an evolution in the world of robotics and represents our primary
focus for the course. Since an operator and a robot work in such proximity, the three
factors, that is, force, speed and path of the robot must be controlled or restricted. Now
that we've established our fundamentals, let's take a look at the robot industry evolution.
We'll also transfer knowledge from industrial robots to collaborative robots.

The Robotics Industries Association has prepared the technical report RIA TR R15.606-2016. They
will enhance the safety of personnel associated with collaborative robot systems. Although
it is a direct adoption of ISO/TS 15066, it is a technical report and has not matured into a
standard yet. The intent is to gather inputs from the standard users, for example, robotic
integrators, which feed into the technical committee. The technical report makes
normative references to various standards. ISO 102018-1:2011, Part one, talks about
safety requirements for robots. Section two, talks about robot systems and integration.
ISO 1200, specifies the fundamental principles to achieve safety in the design of
machinery. Risk assessment and reduction tools are included to continually mitigate risks.
ISO 13850 and ISO 13855, talk about emergency stop functions and positioning of
safeguards with respect to the approach speed of a human body. IEC 6204-1, includes
general requirements for electrical equipment used in machinery. Collaborative robots
and their applications is a continually evolving field. This lesson, we'll build an appreciation
and understanding of different parts of the technical report. This knowledge will be
beneficial in proactively thinking about safety essentials for collaborative robots. Let's dig
in. The technical report takes a process approach to establishing safety essentials for
collaborative robots and systems. There are two central concepts in the technical report,
which are collaborative robots systems design, Clause Section 4 and requirements for
applications, Clause Section 5. Clause Section 4 talks about inherent safety design features
which incorporate requirements for design factors, hazard identification and risk
assessment. It also addresses an approach to mitigate and reduce the effects of a hazard.
Clause Section 5 prescribes requirements for collaborative robot system applications. The
section covers the design and development of work spaces, robot operations and lines of
defense, for example, safety weighted monitored stop. It also includes speed and distance
essentials and preventive hazards by continually assessing and mitigating risks. Clause 6
talks about verification and validation requirements for the design and applications. The
technical report makes a normative reference to ISO 102018-2:2011 Clause 6 on validation
and verification requirements. However, the standard doesn't prescribe how to validate.
My associate Ryan Canlow will be spending some time sharing his insights on a unique
testing methodology a bit later in the course. Clause 7 talks about documented
information required to meet safety objectives successfully. Let's look at potential contact
situations between a robot and a human. However, at this time, let's pause and clarify the
definition of a collaborative robot or cobalt. A cobalt refers to a robot with the capability
of being used in a collaborative operation. A collaborative workspace is a particular kind of
operation in which an operator and a robot share the same workspace. In this type of
operation, the robot is power and force limited making it safer for both the human and the
robot. In this collaborative scenario, a potential contact can occur between an operator
and a robot. These contacts can either be intended, incidental or a failure of the cobalt.
This contact can lead to two scenarios, quasi-static and transient contact. Quasi means
apparent and static means at rest. A quasi-static contact is one where an operator's body
part can clamp between the moving part of the robot and another fixed or moving part of
the robot system. In this scenario, the operator's body part can potentially crush and the
operator is likely unable to retract the body part from the contact. Transient means
momentary, is a type of contact where the operator's body part does not clamp unlike
quasi-static. This momentary contact means the operator can quickly retract or recoil from
the moving part of the robot system. Delving deeper into the possible contact scenarios,
let's review a study performed by the University of Minds. The university conducted a pain
threshold study in Germany across approximately 100 test subjects. The research sampled
pain limits of test subjects by exposing 29 human body parts to robot impact concerning
force and speed. This research yielded the biomechanical limits of the human body in
reaction to robot's force and pressure. Besides, the ISO 102018-1 and 2 talk about
calculations for robot speed to adhere to within the biomechanical limits. Although, this
technical report includes the body model descriptions and biomechanical limits, it's
continually evolving with the ultimate goal of being incorporated into its own standard. If
you recall, we spend some time learning about quasi-static and transient contacts. The
biomechanical limits for the applicable quasi-static and transient contacts for the 29 body
parts were determined. As you can see, in the table extracted from the ISO/TS
15066:2016, which appears as anix A in the technical report, the maximum permissible
pressure and force to specific body areas are calculated for quasi-static and transient
contacts. The top two rows in the table are in italics indicating the skull, forehead, and face
body regions were considered critical zones. We spend some time talking about
collaborative robots, system design, and applications safety. Do you recall the safety
pyramid consisting of the three E's, elimination, engineering and enterprise controls?
Although, the technical report is at a premature stage, it touches upon the two most
 useful competence, elimination of a hazard and putting engineering controls in place to
prevent a hazard. The third level, enterprise price control, reviews how we can proactively
enhance the safety of personnel by changing the way people work. This brings us to OSHA
which stands for Occupational Safety and Health Administration. OSHA's mission is to
improve workers safety by generating new safety knowledge and transferring existing
safety knowledge into new domains such as collaborative robots. Although, not as
prescriptive as ISO standards, there are numerous safety lesson's worth leveraging from
OSHA, such as walking and working surfaces, firefighting, personal protective equipment,
PPE, hazard communications and ergonomics. In this lesson, we reviewed the essentials
and structure of the RIA technical report to stimulate your thought processes and
approaches regarding safety measures.

We learned about collaborator workspace in previous lessons where an operator and a

collaborator robot share the same space and perform tasks concurrently. As humans and
robots begin to work closely, this proximity poses a potential hazard. A hazard is a
condition capable of causing harm or an injury to an operator, or damage to the robotic
system or environment. A hazard is measured on a spectrum by relative severity or
impact. A risk is the possibility of danger that is a degree to which a hazard is likely to
occur. Collaborative applications require a task-based risk assessment. When you're
talking about the old robotic system, you had to have fancy, you had to have guarding, you
had to have safety controls, usually as a fixed box, fixed panel fixed everything. When you
remove the fencing and you simplify the safeties and you get rid of all the hard wiring
going back to the robot controller, now you can make it mobile. But you make it cost
effective too, because all this extra safety that used to be around isn't there. Now, we've
been primarily talking about force feedback robots, pressure force feedback, but there's
also different methods of collaboration. There's hand guidance collaboration, there's
sensors stop collaboration where you get to walk within a boundary of a scattered and the
robot shuts down. You have also the speed monitoring control systems now, where as you
walk into a zone, does a stop at this. It slows down. As you get closer, you get slower and
slower, then stops. But most skills they focus primarily right now on the force and
pressure. When companies are taking this on, back in the days, I used to have to know all
about the safety, about the safety golden rules standards, they have to know about the
whole systems. They had to know safety PLCs or just regular PLCs or control systems
retiring all this wire and be safety relays or whatever. Now, with these newer systems, a
company has to know a few things. One they still have to understand robotic safety. They
have to understand what the safety application now CUDA application. Just go to the arm
is collaborative arm, doesn't mean it's collaborative application. Perfect example, a
company wanted to do grinding. Well, a robot moving a grinding wheel realm can still hurt
somebody. If they grind wheel makes contact with them. A robot move in a box around
has lot less chance of hurting somebody because you don't have that spinning media.
Robot weaving a knife around, although it's collaborative robot, It's not a collaborative
application. So, the companies have to really look at their application say, "Is this truly a
collaborative application, or is it not?" As they get into robotics and advanced robotics,
 they have to understand robotic safety. It's paramount. Because you're still who held two
rules about safe operation of robots, or should be. You have both the RIA standards, the
NCRA, and the ISO standards that need to be looked at. They're becoming more and more
identical to each other as time goes by, but still knows Bob. Therefore, if you're going
collaborative robotics, you have to know about the tactical standards for collaborative
robots, and understanding impact forces and transient energies. When a robot collides,
where you're going to get hit? How you're going to get hit?

Play video starting at 3 minutes 55 seconds and follow transcript3:55

That's collaborative with, in a way, safety is a little simpler with industrial robots. You have offense
and you have a robot, yes you have to understand more about the controls and the
architecture how the tie the safeties in, but it gets a little simpler. So, collaborative for
more effective may be more cost effective. But you have just as much to learn about the
safety especially doing it well that you do as an industrial robot. The value of a risk
assessment is to identify hazards in the work space before it occurs. The risk should be
evaluated for its impact on the operator or enterprise and accordingly implement controls
to mitigate the risks. The value of the preventive approach detects and tackles potential
failure modes downstream before it can cause harm to an operator upstream. The risk
assessment and corrective actions should be frequently implemented, documented, and
updated. Although a facilitator often leads risk assessments, it's a team effort that pulls in
the diversity of thoughts, opinions, and experiences. The brainstorming team should be
comprised of integrators, equipment installers, operators, engineers, and any other
stakeholders. Let's set the context for when to perform a task-based risk assessment.
There are two scenarios. First, conducted a risk assessment for reasonably foreseeable
tasks and associated hazards. Second, when the collaborative robot, its application or its
intended use is modified. First, define the intended use of the collaborative robot. The
scope should include the application, process flow, maturity of the system, personnel with
authorized access, and limits. The second step is to identify the different tasks identified
with the operator and collaborative robot as well as the hazards associated with these
tasks. Repeat these two-step exercise until all the tasks and hazards are identified. Once
the team has brainstormed different risks, the third step is to quantify and qualify the
risks. This assessment with leveraged risk evaluation criteria that we will talk about later in
this lesson. This evaluation allows prioritizing the most severe to the least severe threats.
The enterprise must determine its risk appetite and act upon the high priority risks. Once a
corrective action is implemented, reassess the risk for an increase in its degree of safety.
This loop mechanism allows one to monitor and mitigate risks continually. Document the
whole process. This documentation supports proactive safety engineering instead of the
reactionary afterthought.
A risk assessment is an evaluation of tasks and associated hazards on the basis of injury severity,
exposure, and avoidance. Using a decision matrix, we will determine the risk level of
hazard. Higher the risk level, higher the potential for a task to cause harm or an injury to
an operator. This quantification will prioritize improvement activities. Also, the enterprise
defines its acceptable level of risks. We're going to start with severity. An injury severity
represents the degree of possible harm to an operator. This tie into a hazard associated
with the task between the operator and the collaborative robot. The consequences of the
hazard range from S1 to S3. The S1 rating represents a low severe injury that can be
addressed using four state, and the operator can often immediately return to the job. The
S2 score represents moderate severity. This rating requires medical treatment, and the
employee is likely to recover from the incident. S3 represents high seriousness. This rating
requires medical treatment, and the employee can suffer long-term disability,
hospitalization, et cetera. Moving on to the second factor of risk assessment, an exposure
associated with frequency or duration represents the likelihood of an operator being
exposed to an hazard. The E1 rating represents a low exposure to an occasional activity. E2
represents a high exposure to a routine or a continuous operation. Once a risk has been
evaluated, E0 is an option for verification and validation of risk reduction measures. The
third component of a risk assessment is avoidance. This represents the ability of an
operator to detect and avoid the potential failure mode or hazard. It is measured based on
a function of clearance distance and reaction time. A1 represents a possible avoidance
where there's sufficient clearance to move out of the way and there is an adequate
reaction time. A2 represents insufficient clearance to run out of the way and reduced
speed control is used. A2 rating doesn't have adequate reaction time. A3 represents an
insufficient clearance to move out of the way and reduced control speed is not used. An E3
incident, such as an explosion, has almost no reaction time. The risk level decision matrix is
an excellent tool that captures our discussions thus far. It navigates through the three
essential factors to quantify and qualify the appropriate risk levels. It's important to follow
the order of severity, exposure, and avoidance to derive the right risk level in the matrix. A
high risk level poses a more significant potential for a task to cause harm or injury to an
operator. The risk assessment should yield the risks associated at the task level. Once the
enterprise determines its acceptable risk level, it should implement appropriate risk
reduction methods. Once executed, the risks should be reevaluated to test for a decrease
in risk level or effectiveness of the proactive controlled. The corresponding risk reduction
control based on the risk level shall comply with the table four in RIA TR R15.306-2016. A
risk level of medium, high, or very high requires one or a combination of the risk reduction
measures in the upper quadrant as the primary line of defense. Achieve this compliance by
incorporating an inherent save design in the form of elimination, substitution, limited
interaction, or safeguarding. In addition to the first line of defense, you may use additional
risk reduction measures like warning signs, administrative controls, and personal
protective equipment, PPE. Based on a low or negligible risk level, you can implement any
of the risk reduction measures from elimination all the way to PPE. If a safety-related part
of a control system, SRP or CS, is part of the risk reduction measure, the risk level guides
the minimum functional safety performance. The minimum functional safety performance
is described in ISO 13849-1:2015, safety-related parts of control systems. For a negligible
 risk level, a performance level of b is minimum and can be achieved through an awareness
device like a flashing bulb or a missile. For a high risk level, performance level e, structure
category four is selected, where safety-related functions like emergency stops or control
devices are used.

A robot safety incident can occur for a variety of reasons. There are three broader classifications.
First, a robotic arm or equipment can cause an accident, putting an operator in a risky
situation. Second, a mechanical part fails in the robot system. Lastly, there could be an
uncontrolled power supply to the robot. All of these pose severe risks to an operator. Let's
understand robotic safety incidents and systematically apply them to cobots, minimizing
incidents. Our philosophy here reminds me of a Chinese proverb, "Give a man a fish and
you feed him for a day, teach a man to fish and you feed him for a lifetime." Based on
safety incidents just mentioned, robotic accidents have four broad classifications. Impact
or collision, crushing or trapping, mechanical, and others. We'll spend the next few
minutes understanding the causes and effects of these accidents. Once we gain this
understanding, we'll be well-prepared to apply this knowledge from robots to cobots. Even
if you do get a club robot, people make this assumption that it's safe. It's not the case.
Even with any robotic application, with any robotic system being clever or non clever, you
still have to do a task based risk assessment. You have to figure out every possible way a
human's going to interact with the system and figure out where the potential hazards are,
you walk through them, you think about how you eliminate them or alleviate them. Now,
the club of robots have given me more tools to alleviate these problems. Before, the old
days, task-based reassessment, possible hazard, robot swinging around taking somebody
else that may in it's position. Obvious solution, fence, safety guarding, switch. These days
it's okay, robot swings around hits somebody, evaluate forces of impact. The rules have
changed. So, first and foremost, I always go back to is it truly a collaborative application?
Then even if it is, it,s not, you can't assume it's safe, you still have to go through the steps
to make sure you have a safe system that your people are protective. The first type of
accident is an impact accident. Also known as collision accident. Unpredicted changes or
movement related to the robot equipment or parts such as the peripheral equipment or
competent malfunction can result in contact between a robot and a human. This type is
similar to a car crash which delivers impact to the driver or a passenger inside the car. The
cause is unpredicted moment or component malfunction, and the result is an impact or
collision. The second type of accident is crushing accident. Does this is ring any bell? It's
very similar to quasistatic contact we study. In the scrapping incident, a robot part such as
an arm or equipment can trap between an operator's body part or the operator can be
physically be crushed against another piece of equipment. An example of this accident is a
forklift physically forcing an operator to smash against another standing truck. The cause is
a trapping moment by the robot and the effect is crushing or trapping. In a mechanical
part accident, a robot tooling or any factor such as screwdriver nut fastener etc can break-
down resulting in mechanical failure. This breakdown could lead to multiple hazards.
Mechanical failures can be severe depending on the speed and pressure of the contact.
The cause is the failure of a mechanical part and effects are severe depending on if it
resulted in a contact or crushing. Other accidents can result from working in proximity
with robots. Electrical wiring of the robots and their control power represent a significant
hazard to an operator. A high pressure cutting stream can occur if a hydraulic proline is
ruptured. Environments with radio frequencies, dust, and scrape pose threats. Slip
surfaces in cold regions and excessive wiring present a tripping hazard. The cause of these
accidents are diverse. If you notice, this is precisely where knowledge transfer from OSHA
ties in. As we spoke about in a previous lesson, the effect of these accidents can vary
anywhere from a bruise or a scratch to a life threatening incident. Collaborative robots are
continually evolving field and there's still some discovery to do concerning safety. With so
many variables, the preventive action process is by far the most safest, efficient, and cost
effective practice. Although having a reactive process such as a corrective action or
problem solving is necessary, they take up time, capital, and resources. Our robust
preventive action on the other hand focuses on stopping an undesirable event from
happening in the first place. We have a good understanding of different types of robot
accidents. Now, let's understand why an accident occurs and try to prevent it from
happening. We will identify the root cause of a problem and remove it so that the incident
is likely never to occur in the first place. We're applying our robot lessons to cobot. Here is
an extensive list of identified root causes or sources of hazards in the different type of
accidents. Although one or more of these sources can cause an accident, a good starting
point is to extinguish these hazards. Let's take a moment to identify and define these root
causes. Human interaction with a robot has the potential to cause dangerous or
unpredicted moment or reaction by a robot. Control errors are intrinsic or inherent faults
or bugs within the control system. These errors could also be in the soft fair of the robot
system. Unauthorized access is when personal are unfamiliar with the safeguards in place
or their activation status. A mechanical failure like we spoke about, is conducive to an
unexpected operation or end-effector failure. Power systems pose serious fire risks costs
by disruptive electrical signals. Lastly, the design of a collaborative robot cell is critical. If
the design or the layout is inadequate, it can lead to inherent risks in the system. These
root causes can exist in silos or can be combined to cause an undesired event to occur.
Hence the proactive preventive approach. Identifying root causes is the first step to
realizing a safe workplace. I encourage you to brainstorm with your diverse team to
identify further sources of hazards and how you can prevent them. Putting it all together,
we learned different kinds of robotic accidents. We now have an understanding of their
cause and effect. We problem solved and identified the root causes of these accidents.
Being proactive preventive advocates, we looked at the sources of these hazards in the
context of cobots and discussed how to extinguish them. Now, if you connect this tool to
the risk assessment, you will be able to identify your immediate action items based on
your risk priorities and present the hazards. We not only fished but also gathered an
appreciation for learning to fish. The beauty of this philosophy lies in its flexibility. We're
certain when you apply the proactive preventive approach to your organization or cobot, it
will reap an efficient, beneficial, and ultimately safer workplace.
A quick refresher from our previous lesson on our systemic preventive approach. We gathered in
appreciation and understanding of different types of accidents, their causes, and reverse
engineered to extinguish these sources. We also learned about the three e-pyramid of
elimination engineering controls and enterprise controls which measures the degree of
safety. The purpose of this lesson is to gain proficiency in applying these tools to two-case
studies. We will identify the nature of these accidents in the case studies, their root
causes, and preventative techniques. Then we'll measure the degree of relative
effectiveness of actions using the safety pyramid. These methods, when applied, should
result in a safer, cobot workplace. Case study one. A 23-year-old carousel operator at a
meat packing plant was killed when his foot tripped a light sensor causing a computer-
controlled robotic platform to come down from above crushing his skull. Number one, it's
a crushing accident. He had been watching a maintenance technician work on a conveyor,
and apparently stepped on the conveyor to get a better view of what the mechanic was
doing. The conveyor, the mechanic was working on had been shut off but the entire
system had not been locked out. Power still remained to the light sensors and the robotic
platform. When the platform came down, it pinned the victim between the platform and
the conveyor which confirms a crushing or collision accident. When they lifted the
platform off the victim, the victim fell approximately 15 feet to the concrete floor below.
That was most likely caused by the blow to the head from the robotic platform, not the
fall. The victim was pronounced dead at the scene. Take a minute to reflect on this
incident and how to prevent this from happening in the future. An investigator at the
Nebraska Department of Labor investigated the incident and concluded the following
steps to prevent similar occurrences. The investigator suggested that before maintenance
all equipment is properly locked out or tagged out. A second suggestion included
implementing on the spot inspection to verify employee compliance with safety
requirements. Third, he suggested standard operating procedures to ensure that the non-
maintenance personnel are not in immediate proximity to the maintenance area while
maintenance is active. Lastly, he proposed installing a protective gate around the robot
operation area. As we did in the previous exercise, extinguishing these root causes would
prevent such an incident from happening again. Now, let's take a look at another case
study. On January 19th, 2001, a 29-year-old male died from injuries sustained while he
was stuck on the head by a cycling single-side gantry robot. This also sounds like an impact
or crushing exome, let's keep reading. The victim had recently performed a mold change
on a 1,500 tonne, horizontal injection molding machine, HIM. He was apparently looking
for tools that he may have left within the machine during the setup operation. Let's play
closer attention here. The victim climbed on top of the purge guard and leaned over the
top of the stationary plate in of him in an attempt to see if the tools were left within the
mold area, and place this head beneath the robot's gantry frame. His position placed him
between the robot's home position and the robot's support frame on the stationary
platen. While trying to look inside the mold area, the robot cycled. The victim's head was
struck from the side and crushed between the robot and the robot's support frame. Take a
minute to reflect on this incident and how to prevent this from happening in the future.
The Nebraska Department of Labor investigated this incident as well. The investigator
concluded the following steps to prevent similar occurrences. The investigator suggested
 the robot and its workplace be safeguarded to prevent entry while the robot is active.
Second, the users of the robot system conduct a risk assessment of the robot or robot
systems to identify the peripheral equipment, standards, and task-based hazards. The risk
assessment outputs should provide employee safeguards. Third, personnel who interact
with the robot for example operator, maintenance personnel, programming engineers,
train on safety associated with the robot at the task, and robot system level. As we did in
the previous exercise, extinguishing these root causes and implementing them would
prevent such an incident from happening again.

As robots come into closer contact with people, we need to understand how to design safe
workcells, and how to test collaborative robots, and how to design safe robots. In this example, a
robot assists a hair transplant surgeon in removing and relocating plugs of tissue that contain hair
follicles from one part of the scalp with dance hair to another location with less hair. This involves
the use of sharp surgical instruments and direct contact with a person. In the earlier sessions in
which we discuss risk assessment, this would present a high level of risk, and requires a category
three robot with performance level D. Which requires that no single failure in the robot can cause
an unexpected motion, and that all safety systems must be tested and confirmed before enabling
the robot motors. This level of redundancy in the robot design reduces the risk of the robot
making an unplanned motion. However, in this example, there is application software controlling
the robot, a machine vision system, an operator interacting with the application software, a
patient who may move or recoil, a sharp tool, and possible external disturbances such as a power
failure. Creating safe workcells involves far more than just using a safe robot. In this lesson, we will
discuss techniques for designing safe workcells. Be it collaborative or standard industrial it's pretty
much the same steps: Consider your reach, consider repeatability. How accurately do you have to
pick it up? How accurately do you have to set it down? If you're not pick it in place, then how
accurately do you have to position whatever you're doing, be it a ceiling, be it a grinding, be it a
polishing, be it anything. How accurately do you have to do that? Over what space do you have to
do? What's the weight of your parts? How repeatable is the part you're working with? Do you
need to augment your process with sensors? What's the space you're working in? What's your
constraints in terms of barriers that you're working around or machines you're reaching into? How
are you going to talk to the machines? Or how are you going to talk to the processes around you?
How are you going to do that methods? So, these are all things you think about. Are you going to
be working truly collaborative where you have a shared operator space? Robot operator working
next to each other? Is it robot leaves and operator comes in? These are things that need to be
considered because that affects your safety, maybe operation and maybe robot selection as well.
How fast is your production rate? How are you going to get it from point A to point B? We haven't
even talked or even touched upon end-effector selection, or feedback, or we've touched briefly on
inspection. But there many things that go into figuring out how to successfully build and integrate
robotics applications. But if you're talking specifically about cobots, it's one, are you within the
payload of the cobot? Are you within the reach? Are you within the repeatability? Because not all
cobots are the same. Some have various levels of repeatability. Get into the details. We have
mentioned earlier in this course, the International Standards Organization has developed two
standards that apply to collaborative robots: ISO 10218 and ISO/TS 15066. These standards
defined several different types of collaborative operation. In a monitored stop workcell, a laser
scanner or other sensor monitors the collaborative workspace. In this example, the collaborative
workspace is divided into four different safety zones. There are actually two robots in this workcell
shown in orange. When the operator moves into zone one, the robot stops and stays in zone three
while he loads the tire onto a fixture. When he moves out of zone one, the robot picks up the tire
from the fixture, and loads the machine. Zone two and four work in a similar manner. The laser
scanner is mounted between the two machines, and software allows the scanned volume to be
separated into zones. In this example, it is likely that the laser scanner is actually communicating
with a supervisory programmable logic controller, which is coordinating the two robots entire
machines. Note that, safety mats or other sensing methods could easily be used here. But also
note that the robot motor power stays on while the operator is entering the collaborative
workspace. Well, here's an example of hand guiding. In this example, the robot picks up a 30
kilogram tire from a palette, and places it in the trunk of an automobile. The operator can teach
the robot by pushing on it, and the robot will stop if it bumps into the operator or any other
obstacle such as the car body. In this example, hand guiding is mostly used for teaching and fine
tuning the top positions, or pushing the robot out of the way if there's a collision. This large robot
uses a six-degree of freedom force sensor, located in a stand below the robot, which detects both
operator and collision forces. There are other examples of large robots used as cobots where the
robot is supporting a large payload, and the operator pushes on the payload to guide it into place.
In these examples, there is typically a force sensor or sensors that can detect the non-gravity
forces on the robot wrist that are applied by the operator pushing on the payload. Many smaller
robots with low payload ratings support a gravity compensated hand guiding mode for quick and
easy teaching of motion destinations in the workcell.

Designing Safe Work Cells, Part 2. >> In this picture, we can see an example of speed and
separation monitoring. There are two laser-guided distance sensors located at the base of the
robot stand. As a person moves closer to the robot, the robot slows down. The robots speed will
be determined by how far away the person is from the robot safe zone and by the distance it takes
the robot to stop. In this example, once a person gets within one meter of the robot, it stops all
together. In this case, the robot itself does not sense the distance to the person and does not
sense any collision with person. The person is excluded from robot's workspace by the distance
sensors at the base of the robot. So it could be argued that this example is not really collaborative
as the person and robot can never move in the same zone at the same time and the robot should
never contact the person. Here are three examples with power and force limiting through inherent
design or control. Each of these robots has a nominal payload of 500 grams and cannot apply
enough force to injure a person with a typical end effector that has rounded edges. It should
always be noted, though, that attaching a dangerous end effector, for example, a needle, will
mean the work cell will need additional safety screening to prevent possible injury to a person.
These robots can easily be hand-guided to teach positions without fear of operator injury. They
can be intermixed with people on the same manufacturing line, allowing people to do more
complex operations, and the robot to perform simple operations. These robots can typically move
at speeds up to 1.5 meters per second and bump into an operator without the injuring the person.
In this video, we see a low power collaborative robot that is inherently safe by design. An operator
can even put his hand into the pinch point of this robot and the robot will stop automatically,
without injuring the operator. This is a much faster response than trying to find and press an eStop
button. In addition, there is work going on now that explores the use of capacitive sensors to
detect people from a distance and slow down or stop a robot. Most of these sensors work up to a
maximum distance of about 200 millimeters. They are likely to be most useful for relatively low
payload robots moving at moderate speeds of less than one meter per second. They require
antennas, hidden under the robot cover, which can sense distance and the direction of motion.
This class of technology is being developed for hand gesture sensing in automotive applications,
and as a result, may become cost-effective and reliable in a few years. ISO 10218-2 and ISO/TS
15066, both address the design of safe work cells where the robot has additional application
tooling installed. Even a safe robot can be dangerous with a dangerous end effector or loading a
dangerous process. For example, a stamping machine. What may be less obvious is potential
danger from the motion path and the speeds of the robot through the work cell. As will be
described later in the session on testing, it takes much less force to stop a collaborative robot in
free space where there is room for an operator to slow down the robot over some distance. Than
to stop a robot when it traps an operator against a rigid surface. Consequently, work cells should
be designed, and the robot motion path within the work cells programmed, with the robot speed
adjusted for the distance to the nearest rigid surface. The farther away the robot is from a rigid
surface, the faster it can go, up to free space speeds of one to two meters per second. However, as
a robot approaches a rigid surface, it should be programmed to slow down so that if it traps an
operator against the rigid surface, the forces will not exceed the allowable standards. In the
example in the graphic, the 100% joint speeds of the robot are set to values that allow higher
speed motions. For example, one meter per second, while the 100% Cartesian speeds are set to a
much lower value, for example, one tenth of a meter per second. When the robot is inserting parts
into fixtures, the Cartesian mode is used and this greatly reduces the speeds for the last 100
millimeters of motion down into the fixture. In this example, the fixture approach distances are set
to 100mm above the fixture so that the higher speed joint motions can not trap the operator
against the fixture. It is also important that any controls necessary to operate or stop the robot in
a safe manner be obvious and simple to use or built into the collaborative operation of the robot.
Historically, work cell designers would place one or more large red eStop buttons in robot work
cells. However, when operators are inside the work cell and perhaps even guiding the robot
around by hand, work cell eStop buttons may not be easily accessible to the operator. So the
operator should be able to stop the robot by just pushing on it. It is also important to consider any
cases where the robot may move while the operator is inside the collaborative workspace and
may be startled by the robot. It is a good idea to avoid sharp edges in all parts of the workspace
and to protect an operator who may be startled. [MUSIC]

Here's an example of a simple collaborative robot work cell, which does not require any safety
barriers and is open and easily accessible by operators from several directions. In this cell, a force
limited robot with a nominal payload of 500 grams, loads an instrument with trays picked from
racks. The robot is trained by moving it around in free space and a gravity compensated free
mode. The cell process sequence is managed by a PC, which talks to the robot and instrument via
Ethernet. For this cell, the layout is influenced by the need to have operator switch out the racks
of trays, which can be removed from the outside. Even if an operator reaches into the cell and the
robot bumps the operator or traps the operator's hand against a surface, the robot force is so low
that the operator will not be injured. This type of simple cell can be set up and programmed in a
few days. In this fairly extreme example of a collaborative robot, a good deal of engineering was
performed by the robot designers to provide a high level of redundancy and checking, including
redundant speed checking to ensure the basic robot is safe. The collaborative volume is defined as
a local region on the patient's scalp. When moving inside this collaborative volume, the robot
speed is restricted to a few millimeters per second. There's a force sensor in the tool that causes
the tool to retract and the robot to move away if the force threshold is exceeded. There is also a
motion sensor to detect that the patient's head moves causing the robot to move away. While a
machine vision systems selects hair follicles for harvesting, the doctor reviews each selection and
can modify the harvesting position and angle if desired. The vision system works at 60 frames per
second, and the robot follows any small motions of the patient. A great deal of design effort and
testing went into developing the user interface to make it intuitive and easy to use. Doctors can be
trained on the system in 90 minutes. About 350 patient trials were performed in order to obtain
FDA approval for this system for hair harvesting. In 2018, the FDA approved the system for hair
implantation. Very few robot applications are sophisticated as hair transplantation. Most
applications for collaborative robots are much simpler, such as loading a test fixture, or picking
and placing a component in an assembly. All applications should employ a similar approach to
work cell design. First, perform a risk assessment for the system based on the force the robot can
apply, the type of application tooling, and the motion speeds in the workspace. The workspace
should be designed to minimize pinch points, and eliminate sharp edges. The motion paths
through the workspace should slow down when the robot or tool gets near pinch points. The exact
speed limits will depend on collision forces for the selected robot with the tool and payload. Some
manufacturers publish these forces. For others, testing of collision forces may be required. For
larger robots, the robot may need to stop moving as the operator gets close. Once the robot has
been selected and the motions programmed, the application should be tested for foreseeable
errors, including sensor errors, human errors, and foreseeable external factors such as power
failure. Safe work cell design become second nature after a few designs. However, designers
should record their design criteria as safety agencies or safety officers may wish to review these
criteria before approving a work cell.

In this lesson, we will get into a bit more technical detail on factors that influence the collision
forces for a collaborative Robot. Some of these factors such as mass and velocity are fairly
obvious. However, some other factors such as reflected inertia and the control system design are
not so obvious. ISO 15066 refers to effective payload in several places. We will explain the items
that contribute to effective payload. Also, mechanical design choices such as gear ratios can have a
large effect on collision forces and certain kinds of speed reducers are not very well suited to
collaborative robots. We have described the use of area or region sensors to slow down or stop
robots when people enter the collaborative working volume of larger robots. But for robots that
fall under the category of power or forced limiting by inherent design or control, the robot control
system can have a large effect on collision forces, and we will give an example of how this works.
There are three major factors that determine collision forces. These are: the motor torque applied
during the collision, the kinetic energy of the robot, and the distance to decelerate. For many
robots during a collision the motor torque will spike to its maximum as the path tracking error
suddenly increases. However, it is possible to limit motor torque during a collision as we will
explain later. The kinetic energy of the robot and its payload is defined as one half of the effective
mass times the square of the velocity. Note the energy goes up with the square of the velocity, so
doubling the velocity means four times the collision energy. Finally, the distance to decelerate has
a major influence on collision force. Airbags had been added to automobiles that provide a
cushion of several feet to allow people to survive collisions at highway speeds that otherwise
would have killed them. In a similar manner, adding foam to a robot structure or to a rigid surface
in a work cell can add many additional millimeters of travel for deceleration, and also spreads
force more evenly over the colliding objects. This can be very helpful for example, in protecting
the human hand, which has a lot of delicate bones. Imagine being hit by a wooden baseball bat or
a foam Nerf bat. The effect of payload includes the actual payload, the mass of the tool, or end
effector attached to the robot to carry the payload, the inertia of the robot computed as a mass at
the radius of impact, and the forward reflected inertia of the motor and drive train computed as a
mass at the radius of impact. In 1983, General Motors performed a study on the weight of all the
parts in an automobile, as part of a program to define the specifications of a robot for assembly.
GM found that 95 percent of the parts in an automobile weigh less than five pounds. Even though
there are a number of large heavy parts in a car, when you consider the electronics, the
instrument panel, the door locks, the brake system, the fuel injectors, the pistons, et cetera. The
vast majority of the parts weigh less than five pounds. However, because small parts were easy for
people to handle, most robots in the automotive industry were developed for large heavy parts. A
lot of small part assemblies were outsourced to low cost labor in Asia. Today, the labor in Asia is
no longer at low cost, and as a result, there is a rapidly increasing interests in using robots to
handle the assembly and testing of small parts in the US, Europe, and also Asia. However, in the US
and Europe, we often still see powerful large robots used to handle small lightweight parts. Not
only is it difficult and expensive to make these large robots collaborative, they waste a
tremendous amount of energy. A person consumes about 100 watts of energy as does a small part
collaborative robot designed to handle a few kilograms. However, a 10 to 20 kilogram payload
robot, consumes 1,000 watts of energy, most of which is dissipated as heat energy. For 100 robots
working in a factory, if a larger robot is selected for a job that a light payload robot can
accomplish, the factory we'll be wasting 90,000 watts in electricity, and will have to remove this
energy with the cooling system. Perhaps the least understood element of effective mass at the
end of the robot is reflected inertia. Reflected inertia, is the inertia of the motor and the high-
speed part of the speed reducer multiplied by the square of the gear ratio as seen at the output
shaft of the speed reducer. The reflected inertia can then be translated to an equivalent mass at
the working radius of the robot where this equivalent mass equals the reflected inertia divided by
the square of the radius of the rotating joint. Here, we compare the reflected inertias for the exact
same motor for two robots, which have dramatically different gear ratios. This table is for the
PF400 robot, which has a timing belt ratio of five to one for a 200 watt motor. The square of this
ratio is 25 and the equivalent mass at a 400 millimeter radius is an insignificant three grams. Here,
is a table for a small harmonic drive robot. Their harmonic drive gear ratio is 160 to one and the
square of this ratio is 25,600. For the exact same 200-watt motor, the equivalent mass at 400
millimeters is 5.9 kilograms, which is 2,600 times the equivalent mass for the low ratio robot. Note
this is just for one joint and typically two or more joints will be involved in a motion. This reflected
inertia equivalent mass at the working radius becomes part of the effective mass and a collision.
So, harmonic drive robots will tend to have much larger collision forces than low ratio robots.

For a given stopping force, the effective mass of the robot will determine the required stopping
distance. If we pick a stopping force of 150-newtons, often used as a rule of thumb for
collaborative robots, we can compare the required stopping distance at several velocities for a
high ratio robot and a low ratio robot. This table shows the stopping distance for two robots. A
high ratio Harmonic Drive SCARA and a low ratio belt drive SCARA, for both a 500 gram and a two-
kilogram payload. At a speed of 500 millimeters per second, the harmonic drive SCARA it takes 12-
13 millimeters to stop while the belt drive SCARA takes one to three millimeters to stop. Recall
that the human hand has a compression constant of 75 newtons per millimeter so that you can
only compress the human hand two millimeters before reaching the 150-newton pain threshold.
Therefore, unless the harmonic drive robot has 10 millimeters or more of foam on the part that
strikes a human hand, at 500 millimeters per second, it will likely cause injury while the belt drive
robot will not. At a speed of 1,000 millimeters per second, the energy is increased by a factor of
400 percent and the stopping distance for the harmonic drive robot increases to about 50
millimeters. Well, the belt drive stopping distance increases to 5-10 millimeters depending on the
payload. Collision force tables, if available, can be used to set speed limits for various types of
robots in various motions. In general, low ratio robots will be able to incur safe collisions at higher
speeds than high ratio robots. In addition to the inertial forces that contribute to the collision
force, we have forces from the robot motors. In many robot controllers, when a collision occurs,
the path following error increases quickly and the motor torque quickly ramps up to the maximum
possible motor torque before shutting down with an error. However, it is possible to design a
control system which limits motor torque in the event of a collision. This can be done by
developing a complete dynamic model of the robot which computes in real-time the motor
torques necessary to make a commanded motion including all the torques coupled back from the
other robot axes. This computation is complex but with today's microprocessors, it can be done
very quickly. The precomputed torque from this dynamic model is sent to the controller as a feed-
forward torque along with the commanded motion. The controller combines this feedforward
torque with feedback torque which is computed based on the path error from the commanded
motion. If the dynamic model were perfect, there would be zero path error and the feedback
torque would be zero. However, since there are always some unmodeled errors such as friction or
tooth cogging in the drive train, there will always be some feedback error torque. Low ratio robots
tend to have low friction. In these robots, the feedback error torque can be as low as 10 percent of
the total motor torque. It is, therefore, possible to set a maximum limit on the feedback error
torque, that is, for example, 25 percent of the maximum motor torque. In a collision, this 25
percent motor torque limit is all that is allowed before a shutdown as opposed to a 100 percent
limit without this feature. In this image, you can see an example of this control strategy. The green
triangular line is the dynamic feedforward computed torque from the robot model. The blue
wiggly line is the feedback torque from the path following error. You can see that the magnitude
of the feedback torque is much lower than the feedforward torque. The yellow line, superimposed
over the green triangular line, is the total torque output from the controller to the motor which is
composed of the feedback torque added to the feedforward torque. When the motor is following
a normal commanded trajectory, most of the torque comes from the feedforward model and only
a small percentage of torque comes from the feedback model. Therefore, it is possible to set a
limit on the feedback torque that is about 25 percent of the total motor torque. A collision will
quickly cause the feedback torque to rise up to this limit in which case the robot will stop. This
limits the collision force from the robot motors to 25 percent of what this force would otherwise
be. Note that this strategy is most helpful for the lower speed collisions. For high-speed collisions,
the inertial forces will dominate and factors in the mechanical design of the robot such as
reflected inertia are most important. Here's a video of a collaborative robot doing a 600-millimeter
pick and place in a few seconds with a one-kilogram payload. An operator can literally place his
hand in the pinch point of this robot and it will stop without injuring the operator. Note that this
collaborative capability is also very useful when the robot is loading an expensive instrument or
test fixture. If the robot bumps into the instrument, it does not damage it in the way of more
powerful robot might. Here we see a small conveyor with two collaborative robots attached to it.
One robot is a SCARA type robot sitting next to the conveyor and the second robot is a
collaborative of Cartesian robot which is mounted directly over the conveyor on standoffs These
are both low ratio robots with very low collision forces. The low ratios also make it easy to back
drive these robots for the purposes of teaching. Let's put this robot into teaching or free mode in
which the gravity load is counterbalanced by the motors and we could move the robot into various
positions with just a few grams of force. This makes it extremely fast and easy to teach the robot,
as you can see. You don't have to fight with high ratio of gear trains to position the robot
accurately. Another benefit of the back probability of these robots is the ease of compliant
insertion. If parts or nests have a modest chamfer on them, robots that are back-drivable can
easily comply in x, y and Theta to allow the insertion to be completed without the jamming that
often occurs with robots with high ratio, very stiff, gear trains. Now, let's start at this line and see
these robots in action. I can walk up and easily stop the robot without any large force.

Play video starting at 7 minutes 44 seconds and follow transcript7:44

In this case, the robot has been programmed to remember what it was trying to do. Even though it
was interrupted, it pauses for a moment and then resumes operation and completes its task. Let's
look at the pinch point on this SCARA robot. I can place my fingers directly in the pinch point and
the robot stops easily without injuring my hand. Then, after a pause, presumes its operation. The
ability to bump a person without injuring the person and then recover and continue performing its
task allows this class of collaborative robots to be intermixed with people on the same assembly
line. These robots are moving at speeds which do not scare people. If they bump the robot, the
robot stops and the person is not injured. Then, the robot can recover without the need to restart
the whole line from an emergency stop shutdown. The ability to play is friendly, collaborative
robots in convenient places in existing lines greatly reduces the time and expense to introduce
robots into the simple applications where they perform best. People can continue to perform
more complex applications and the robots can do the boring jobs.
The much we're hearing live in media today is robotic systems replace jobs or cause people to lose
work. Let's take a case study. Look at the West Coast back in, I'm going to give my time periods
drawn, let's say, look at the West Coast back in the 30's and 40's, their everything was manually
done. They had thousands and thousands of people, but over a period of time they started
automating with cranes, and these large automated cranes were doing unloading. But, what
happened is they allowed our shipping industry to grow, they allowed the industry to grow much
more because we could do more work and take more work, so more products could come in. You
didn't have simple constraints. It's the same thing with automation robotics. So, whenever we look
at robotics, we always focus on the 3D's; dull, dangerous, and dirty. I usually like throw a fourth D
in there, demanding, especially when we talk ergonomics. There's a lot of jobs people just don't
want to do, really they just hate doing that could cause them ergonomic issues. Somebody with a
electric screwdriver in a hand putting the screws all day twisting the wrists, these are jobs that
over time will cause medical issues, these are good advocates for robotic systems, there's
applications for robotics don't have necessarily the entire skill set or effect for affordable enough
to do the entire main assembly process, where you still need these labor to work with the robot to
complete assemblies. We have a local company here who assembles electronics,

Play video starting at 1 minute 46 seconds and follow transcript1:46

they understood that the robot couldn't do everything, and the people could be assisted
augmenting them in a way. In all honesty, there may be some point in the future, a couple of jobs
here and there they'll replace but then the company is more productive, it grows, it hires more
people, because it can be more productive, it can meet the world demand, they can meet the cost
of competitiveness with their competition. So robot are technology for growth, and let's not
dismiss them, it's not the end of the world. There's other examples, take a look at the welding
workforce, year after year after year of attending a welding seminars, the American welding
society has said same thing, "The welding workforce is dwindling, nobody wants to do welding
anymore, it's just non acceptable tests your skill that people want to take on." So what do you do
about that? How do you, things still need to be welded together and they're still a lot of parts
need producing, as those parts are increasing as our human race expands. So, what do you do?
Well, you have to have a way to fill in that need, where there are shortages of people, where you
have shortages of certain skilled labor you have to have a need to back fill them. Then now we're
not making just people skilled labor, we are actually elevating them, so now they're not just doing
welding but they're running a robotic welding system, where you don't necessarily need the skilled
labor but you're also educating them, you making them more capable to work in today's
manufacturing environment. Most people perform tasks at speeds ranging from 500 millimeters
per second to about 1.5 meters per second. If collaborative robots are going to be useful they
need to be able to move at similar speeds. Replacing a person with a robot that can only work half
as quickly is not attractive. Here you see a 4-year-old junior robot tester stopping one of our
collaborative robots from a speed of 1.5 meters per second, the robot stops in a few millimeters
and the tester is not harmed in any way. This demonstrates how good mechanical design
combined with sophisticated control algorithms can allow collaborative robots to move at speeds
equivalent to people while still being safe if they bump into people. To summarize this lesson, we
learned how to minimize the moving mass of both the robot and the end effector through
selecting a robot that is appropriately sized for the payload and designing lightweight end
effectors for small payloads. How to determine when to consider a robot with low ratio drives
rather than high ratio drives in order to minimize reflected inertia, which can add many kilograms
of effective mass at the working radius of the robot. We learned the value of control algorithms
that limit the motor torque in the event of a collision, we learned to minimize sharp edges and add
compliance where possible on robots end efectors and fixtures.

In this lesson, we will show collaborative robots working in several different industries, completing
several different applications. One of the most common uses for robots is in material handling,
ranging from very light payloads to very heavy payloads. To date, collaborative robots have been
used in material handling applications with payloads of up to 30 kilograms. Although at heavier
payloads, they are restricted to moving slowly. Most applications are for payloads in the 1-5
kilogram range. The automotive industry is one of the most heavily automated industries with a
lot of heavy, dangerous jobs such as spot welding, spray painting, and palletizing performed by
large powerful robots. The automotive industry has also been quick to embrace collaborative
robots for lower payload applications including some in which the robots will work together with
people. Robots continue to find new applications. Recently, the life sciences industry has started
using robots to automate testing and drug discovery. There are now thousands of robots working
in laboratories around the world. The electronics industry has automated the assembly of printed
circuit boards and now, robots are beginning to be used to load testers and sort tested boards.
Robots are also assembling consumer electronics products such as smartphones and disk drives.
Finally, we will show an example of how collaborative robots can reduce floor space in factories by
mounting directly onto existing conveyors and eliminating large control cabinets and safety
screens. Earlier, we spoke about zone sensors which monitor an area by sensing when people
enter the area and slow down the robot if a person gets too close. In the following video, you can
see an example of this approach to converting an industrial robot to a collaborative robot. This
collaborative robot can handle a payload of up to 30 kilograms. There's a large force sensor
between the base of the robot and the floor. This force sensor records the normal forces during a
training session. Then, if during operation, an unexpected force is detected, the robot will stop.
The force sensor and controller also support a mode where the operator can push the robot
around to teach different positions. As payloads get heavier and robots get bigger, they must
move more slowly, down to speeds of a few hundred millimeters per second in order to work in a
collaborative mode. This is an example of a person and a robot working together as a team to
install shock absorbers in the wheel well of a car. The person loads the shock absorber onto the
robot and the robot reaches up under the car and drive screws into place to attach the shock
absorber. This team approach greatly reduces the stress and fatigue for the person, but avoids the
fairly difficult process for the robot of picking up the shock absorber and inserting the screws into
the bracket prior to installation. In this transmission manufacturer, a collaborative robot was
recently installed to load parts into an automobile transmission. In the video, you will see a worker
delivering trays of parts to the robot. You can see the robot stop if it bumps into the worker, no
safety screens are required. In this video taken at Volkswagen, a collaborative robot aids a worker
in assembling an electric motor armature. While a second robot does some material handling,
loading and unloading the armatures from a conveyor.

Play video starting at 3 minutes 55 seconds and follow transcript3:55

Here, we have a work cell for pharmaceutical research in the life sciences industry. In this work
cell, the instruments can cost hundreds of thousands of dollars. So, not only is there a benefit in
protecting the technicians who load the work cell, but if the robot bumps into an instrument, it
does not damage the instrument. Note the kinematic configuration of this robot with a horizontal
arm and vertical tower which allows it to take up a very small footprint in the work cell, it is still
able to reach into, around, and over the various surfaces and trays in the work cell. The robot
makes a number of complex motions to orient and reorient the sample trays as they move around
the work cell. Circuit board testing is a common application in the electronics industry. Many
circuit boards are made in high volumes for consumer products. The assembly of these circuit
boards has been largely automated. However, circuit board testing uses a variety of fixtures as
well as testers, most are loaded manually. In this video, you see a collaborative robot with a vision
system and a gripper that can locate a circuit board, pick it up, place it carefully in a tester, and
then close the lid on the tester. When the test is complete, the robot opens the tester, picks up
the circuit board and sorts it based on the test results. In this factory, all the testers were clamshell
testers designed for manual operators. However, the robot is able to open and close the tester
just as a person would. Here's another life science application in which two robots both mounted
on linear rail access a large working volume. The cell has storage on one side and racks of
expensive test instruments on the opposite side. There is an access aisle down the middle of the
cell to allow people to service the equipment. In this case, collaborative robots were chosen
primarily because of the cost of the test instruments. Industrial robots, which could access this
large volume would be way too powerful if there was a collision with an instrument. Note also that
in this application, the robots hand off trays between the two work cells and must avoid working
in the same area at the same time. Note also the large vertical travel of these robots results in a
very small footprint. One of the characteristics of these particular collaborative robots is that they
have the controls mounted inside the robots. There aren't any external cabinets full of controls
and cables. This allows the robots to be attached directly to conveyors and even mounted over the
conveyors, thus, greatly reducing floor space, as opposed to a larger floor mounted robot with a
large cabinet full of controls. In this video, we see a pallet conveyor with zone control. The
conveyor is divided into 750 millimeter zones and pallets that can move from zone to zone. The
pallets can be stopped and elevated so that other pallets can continue underneath them. This
allows a synchronous timing of operations on the pallets without the need to remove the pallets
from the conveyor line. For example, if one operation takes 10 second and the next operation
takes five second, pallet stations can be assigned so that two can handle the 10 second operation
and one can handle the five second operation. This avoids pulling the pallets off the line and
creating extra work cells for the longer operation. As factory floor space is very expensive, the
ability to eliminate large workstations on either side of the conveyor is very valuable, which is
made possible by compact collaborative robots. This ability to design so-called lean automation
work cells is demonstrated by a comparison of a traditional, electronics, robotics assembly line
with a lean assembly line with a pallet conveyor with clip on collaborative cartesian robots. The
traditional line has large cabinets around each robot in which the bottom of each cabinet is filled
with controls and operators are protected from the robots by expensive enclosures. The lean line
with collaborative robots eliminates the bulky cabinets and the controls, in this case, are located
inside the robots. The lean approach greatly reduces factory floor space and cost. Different
processes can be mounted on the collaborative cartesian robot such as screw driving, dispensing,
assembly, and loading testers. In addition, people could be interposed among the collaborative
robots. For many products, there are some very difficult tasks, for example, wiring. With a
collaborative robot approach, people considered stations to perform demanding tasks and the
robots can be placed next to the people to perform simple and boring tasks. The intermixing of
robots and people means it is not necessary to automate the entire line, which greatly reduces the
time, engineering, expense, and risk that comes with installing robots. In conclusion, collaborative
robots are quickly gaining acceptance in a number of different industries. They are reducing the
cost and risk of installing robots. Collaborative robots are often viewed as friendly machine by the
people who work with them. In many factories, they are given names and considered co-workers.

[MUSIC] Great application for collaborative robots, machine tending. There's company after
company after company that have see the C machines that require a part to be put in, a part to be
taken out. Most of these times these parts are small parts. Collaborative robots are best designed
for weights 10 kilograms and lower. There are larger kilograms arm, up to 35, and there are some
collaborative systems getting up to 100 kilogram. But they use different safetys to make sure they
are safe, but material handling is the best. Any metal fastening processes, industrial robot. There's
some polishing where you're using a soft wheel but even that can be argued depending on speeds
of the polishing wheel, the fluffy wheel. So material handling for collaborative, some sealing
processes maybe, but you have to be careful when you're dispensing tool. It's a nice sharp pointy
needle. That's not necessarily the best collaborative. Some assembly applications are excellent for
collaborative. But once again, you have to pay attention to what's on the end of the arm, what the
application is. But industrial bots are good for 15 kg and higher as the standard for capacities, with
the exception of one arm in the market. Good for high speed. Collaborative robots, due to the
nature of colliding with people, tend to run at a lower speed. So if you have a high speed
application doing 15, 20 picks a minute, that's probably going to be an industrial arm. So, lower
speeds, lower weights are good for collaborative applications. Another area that collaborative
robots have been good for has been low mix. Sorry, high mix, low volume, with the ability to hand
guide and hand teach.

Play video starting at 1 minute 59 seconds and follow transcript1:59

Once you're used to collaborative robots it could be faster to reteach parts or to teach new parts,
allowing quicker changeover. Small areas, and we see collaborative being used as small areas in
plants. So you have a row of machining centers where you have maybe 36 inches between
machines. To put an industrial arm in there and to fence it properly and to make it safe is difficult.
But with the right end of arm tool make sure that the application meets its needs and the safety
requirements for impact forces. You can slide a collaborative arm in there and it'll be fine. So there
is a really good spot. A lot of small and medium manufacturers are, still have simple material
[INAUDIBLE] processes that are being done. Another example, low speed safety processes, say,
moving a basket of parts from anodizing tanks to anodizing tanks. It may not be a high speed
process, because they had time to do the anodizing, but this is something where a collaborative
arm could be helpful. Now this one gets on the fuzzy edge due to the chemicals and the natures.
You still may want to guard that one. But it could be. [MUSIC] So the first thing you do is, well not
everybody does. You almost have to have an eye for it in some ways. But mostly your industrial
engineers can do these walkthroughs and look at your process. It's a matter of looking at your
repetitive processes in the plant, and think to yourself, is there a repetitive motion? Is it a motion
that can be done with an arm? Do you have the proper reach? Companies just need to look at
what they're doing manually, and think to yourself is there a level of artisanship in what they're
doing that's not going to allow a robot to do it easily? If it's somebody taking jars and putting them
in a box, you think immediately, here's an application. If it's somebody loading and unloading a
machine for bending tubes, that's an application. And you walk through and you take a look at it.
But you always have to keep in mind the amount of area space that this process takes. How big are
the parts? How heavy are the parts? How rapid are they coming out? How much do they have
change over? Is it the same part over and over? Repetative process? Or is it a different part every
single cycle? Which makes it a little bit more challenging. A company should always look for easy
wins at the start. Too many times companies have tried to tackle their most demanding processes,
failed and the automation sits for five, ten years before they'd have the nerve to try it again. It's
the easy, quick wins that builds the confidence, gets their feet wet, where they see a benefit. The
may not be as great, or the may be a little further out, but it's more of a confidence builder. And it
gives them the ability to learn.

Play video starting at 5 minutes 17 seconds and follow transcript5:17

There's all kinds of ways of inspecting parts that just come to mind. Let's start with tactile. If you
have force feedback sensors, position feedback sensors, or grippers even, you can measure parts
as you're picking them up if you have force feedback on the other arm. You can measure weights
as you're picking things up, as you're sorting boxes. You have vision systems that are extremely
cost effective that can check dimensionally, that can check pattern matching, that can have
different tools for doing verifications. But you have to be able to tie in these external sensors and
peripherals into the robot so it can grab a part, do an inspection, set it down. So yes, inspections
are possible but the great thing about automating inspections is is that over a period of time the
human tendency is, not necessarily boredom, but fatigue. You don't necessarily get to repeat a
building of 100% inspection. Even though you have this perfect J, over time there's going to be
errors. You reduce the errors greatly when you have a system or automation or machine or vision
doing the exact same way, the exact same time, for the exact same niche of information every
single time. [MUSIC]

What is an acceptable robot modification?

1 / 1 puntos

Modifying robot harness or robot-to-controller cables

Modifying robot access covers or drive system components.

None of the above.

Correcto
Correct!

2.
Pregunta 2
If a robot strikes a person, a severe injury could occur. What factors contribute to the total
amount of energy available to cause injury?

1 / 1 puntos

Robot configuration

Joint speed

Joint orientation

Attached payload

All of the above

Correcto
Correct! All of the factors listed above contribute to the total amount of energy available to
cause injury.

3.
Pregunta 3
Identify the type of robot programming by teaching method: An operator established physical
contact with the robot and drove the robot's arm through the desired positions within the
working envelope.

1 / 1 puntos

Lead-through programming
Walk-through programming

Off-line programming

Correcto
Correct! During walk-through programming, the robot's controller is scanning and storing
coordinates values on a continual basis.

4.
Pregunta 4
Identify the type of path generated: The robot's arm follows a predictable orientation and
trajectory as it navigates in the work envelope.

1 / 1 puntos

Controlled path

Continuous path

Point-to-point path

Correcto
Correct!

5.
Pregunta 5
During normal operation, the operator intruded into the tooling area of the running collaborative
robot system. The cobot's workpiece accidentally clamped the operator's shoulder joint
resulting in a force of 220 N. Calculate the maximum permissible force for the contact and
determine if it is compliant with requirements of RIA Technical Report R15.606-2016.

1 / 1 puntos

For quasi-static contact, the maximum permissible force for sternum is 210 N, and the contact
is compliant with RIA technical report requirements.
For transient contact, the maximum permissible force for shoulder joint is 210 N, and the
contact is compliant with the RIA technical report requirements.

For quasi-static contact, the maximum permissible force for shoulder joint is 210 N, and the
contact is not compliant with RIA technical report requirements.

For transient contact, the maximum permissible force for shoulder joint is 120 N, and the
contact is compliant with the RIA technical report requirements.

Correcto
Correct!

6.
Pregunta 6
You're an engineer tasked with facilitating a task-based risk assessment for a collaborative
robot. During a facilitation session, your team discovered the cobot's screwdriver end effector
threats. The team came up with the following: An hourly exposure per day with inadequate
warning time potentially resulting in a non-reversible accident. Based on this input, estimate
the risk level.

1 / 1 puntos

High risk

Low risk

Moderate risk

Correcto
Correct! Based on the team's input, the severity is an S3, the exposure is an E2, and
avoidance is an A2 resulting in a high-risk level.

7.
Pregunta 7
Based on the above example and collaborative work cell, propose one risk reduction measure
compliant with the RIA TR R15.306-2016 table 4 on minimum risk reduction measures as a
function of the risk level.
0 / 1 puntos
First, you have to eliminate that threat and replace the parts and the robot process. Another
secondary point is to have supplemental security measures such as a fence during
maintenance or procedure.
Incorrecto
Incorrect. For clarification, please see Module 1 Lesson 4 Part 2, which is titled "Risk
Assessment (Part 2)".

8.
Pregunta 8
To temporarily pause the collaborative robot, the operator modified the robot's electrical
component. Is this an acceptable modification?

1 / 1 puntos

No, the modification doesn't affect the cobot and is safe for regular operation.

Yes, since the modification may damage the cobot, reduce system safety and reliability, or
shorten the life of the robot.

Correcto
Correct!

9.
Pregunta 9
Recognize the risk mitigation device in the cobot workcell of the video presented in Module 4.

0 / 1 puntos

Awareness device

Non-mechanical limiting device

Presence sensing device

Incorrecto
Incorrect. Please refer to the video in Module 4.

10.
Pregunta 10
An operator was resting his palm at his workstation when a collaborative robot accidentally
trapped the operator's arm against the workstation. The cobot exerted a pressure of 200
N/cm^2 until the trapping condition alleviated. Is the contact condition compliant with the safety
requirements of the RIA TR R15.606-2016?

1 / 1 puntos

Not sure, the problem doesn't provide enough force and pressure calculations to determine
compliance.

No, the maximum permissible force for a quasi-static contact for palm is 140 N. Hence, is the
condition is not compliant with the RIA technical report requirements.

Yes, the maximum permissible pressure for a quasi-static contact for palm is 260 N/cm^2.
Hence, the condition is complaint with the RIA technical report requirements.

Correcto
Correct!

11.
Pregunta 11
A table mounts a collaborative robot. During its regular operation, a production worker was
facing away from the robot while the cobot accidentally hit the worker's fifth lumbar vertebra.
However, the worker quickly reacted and recoiled from the robot without trapping his contacted
body area. What is the maximum permissible pressure multiplier for this contact condition?

1 / 1 puntos

3.5 Pt

3 Pt

2.5 Pt

2 Pt

Correcto
Correct!

12.
Pregunta 12
A collaborative robot's moving arm impacts an electrician's abdominal muscle during regular
maintenance without trapping the personnel's contacted body area. The contact lasted for a
very short duration. Identify the type of contact and corresponding maximum pressure
multiplier.

1 / 1 puntos

Quasi-static contact and 140 Pt

Transient-contact and 7 Pt

Dynamic impact and and 2 Pt

Correcto
Correct! Transient contact is also referred to as "dynamic impact".

13.
Pregunta 13
An untrained operator reaches into work cell while a collaborative robot is moving and collides
with the operator. Maximum free space collision force is 50 N. In another instance, an
untrained operator reaches into work cell while the robot is moving into instrument tray and
hand is trapped between the robot and instrument tray. Maximum trapping force in downwards
Z direction at 50 mm/ sec is 50 N. Based on the above two scenarios, perform a risk
assessment and derive a PL.

1 / 1 puntos

S1, E1, and A1, and PL = b

S1, E2, and A3, and PL = d

S2, E2, and A2, and PL = e

Correcto
Correct! S is S1 as possible collision forces will not injure operators. E is E1 as normal
operation does not involve collisions with the robot. A is A1 as the robot does not make
unexpected motions. So PL is "b," given the low speeds and small possible collision forces
which cannot injure an operator.

14.
Pregunta 14
Pressure equals force divided by area. A collaborative robot has a sharp end effector with an
area of 0.25 cm^2. During a routine operation, the robot's end effector exerts a 50 N force in
the downwards Z direction on the back of an operator's hand. Analyze the pinch point and
determine if it is compliant with the RIA TR R15.606-2016 requirements.

1 / 1 puntos

A 50 N force in the downwards Z direction and an end effector area of 0.15 cm^2 yields 333.3
N/cm^2 pressure. This contact pressure is way above the maximum permissible pressure of
200 N/cm^2 for the back of the hand and is not compliant with the technical report
requirements.

A 50 N force in the downwards Z direction and an end effector area of 0.15 cm^2 yields 267
N/cm^2 pressure. This contact pressure is way above the maximum permissible pressure of
200 N/cm^2 for the back of the hand and is not compliant with the technical report
requirements.

A 50 N force in the downwards Z direction and an end effector area of 0.15 cm^2 yields 110
N/cm^2 pressure. This contact pressure below the maximum permissible pressure of 200
N/cm^2 for the back of the hand and is compliant with the technical report requirements.

Correcto
Correct!