Collaborative robots are having a profound impact on industrial manufacturing. Robots are taking over the world. Some countries are so concerned with rapid robot expansion that there’s been discussion of a robot employment tax to countermand the displacement of workers caused by robot implementations. Chinese appliance maker, Midea, riding the momentum of its $4 billion acquisition of Kuka, intends to fend off Japanese competitors and dominate the world’s largest robot market.
There’s a lot at stake. With the rise and convergence of the Industrial Internet of Things and collaborative robot applications, the possibilities are endless. And a sure sign of the times was Schunk’s collaborative robot gripper, with embedded intelligence and two cameras, winning the Hermes Award at Hannover Messe 2017.
The sky’s the limit on robotic industrial applications. We asked a variety of industry experts some specific questions about where robots are headed and how they plan to get there.
Industry is abuzz with collaborative robots, but is it the robot or the application that is collaborative? And what is the difference between a robot that can operate uncaged to collaborate with humans and one that can’t? Is there a downside to installing a collaborative robot when the application doesn’t call for one?
Carole Franklin, Robotic Industries Association (RIA): When it comes to collaborative robots, the application is key. Some tasks are simply not well-suited for collaborative operation, even if the robot performing the task is power- and force-limited and is called a collaborative robot. What we call a collaborative robot is simply one designed for use in a collaborative workspace, to perform a collaborative task. But despite how it was designed, its actual use might not be safe for collaborative operation if the workspace has not been carefully planned and integrated for appropriate collaborative operation. The robot does not work in isolation. Remember, the robot arm by itself cannot do any work. The workstation also includes the end effector, the workpiece itself, the potential presence of multiple robots and other equipment in a cell. All these factors and more must be taken into account when companies plan for a safe robotic installation. And this planning needs to take place even—or especially—when companies use robots designed for collaboration. This is precisely because we anticipate people and collaborative robots will be working in close proximity.
It’s also critical for risk assessment to be done from the perspective of the system. Even when using a robot designed for collaborative use, safety standards require companies to complete a risk assessment and mitigate any risks identified in the system.
The best way for people to ensure they are following safe practices is for them to get their arms around the collaborative supplement, TS 15066, to the overall safety standard, ISO 10218. Of course, companies need to remember that the effective use of TS 15066 assumes that the robot system under consideration is in compliance with Part 1 and Part 2 of ISO 10218:2011.
Carole Franklin is director of standards development at Robotic Industries Association (RIA).
Bob Rochelle, Güdel: The collaborative robot is required in the application where collaboration is required. The difference in the uncaged to caged robots is in compliance to the safety-rated requirement for the control system to allow a safe stop. And in some cases the manufacturer uses very low-wattage motors so they stall out when the robot encounters an obstruction. Hence it stops when it hits something. Many of the collaborative robots do not allow the higher speeds of the standard, caged robot, and in some instances these are not as repeatable or precise as the caged design. To use the caged style robot in a collaborative system requires the use of the safety-rated speed and separation technology to allow the robot to slow and stop as the hazard, the human, approaches. The downside to using these uncaged technologies in an application that does not call for it is a sacrifice in speed of operation and precision.
Bob Rochelle is logistics market specialist/area sales manager at Güdel.
Ryan Guthrie, TM Robotics: When considering a collaborative robot, the application is always the deciding factor. Sometimes an end user needs a faster, more accurate standard robot, when compared to the slower speeds of a collaborative robot, which are limited in order to satisfy the safety requirements of a collaborative system. And sometimes a collaborative robot, which can work in close proximity to a human without needing to be caged in, is a better fit. If the final design of a system has the robot working alone, without consistent human interaction, then the designers really need to ask themselves why collaborative is even being considered. There are a number of safety equipment manufactures that can provide equipment to protect a standard robot that needs occasional input from a human and in many cases at a cost savings when compared to the collaborative alternative.
Ryan Guthrie is executive vice president at TM Robotics.
Matthew Bush, Hirebotics: There is a lot of confusion in the industry over collaborative robots, and what they really mean. Even the most collaborative robot when using an end effector or handling a dangerous part still must have safety and security of the personnel working with and around the robot taken into account. In my opinion, it’s the total work package that makes a collaborative solution. Even a robot inside a cage can collaborate with human operators to perform the task at hand. The reason that we primarily deploy collaborative robots even when fully caged is that they tend to be faster to deploy and easier to program, with more built-in features such as built-in force and torque monitoring. Also, we find that they are more versatile, and therefore, when the job changes or goes away, they are much easier to redeploy into other applications.
Matthew Bush is COO and co-founder at Hirebotics.
Maximiliano Falcone, Kawasaki Robotics USA: We receive calls from customers for our new duAro robot. They perceive a collaborative robot as an easy way to get into robotics. In reality, there is in some ways actually a lot more to consider when integrating a collaborative than with its non-collaborative brethren. Within the collaborative standard there are many factors that need to be considered that the robot cannot exceed or be able to detect while running. For this work, these robots run much slower than what a non-collaborative robot would run within a fenced-in cell doing the same application. However, at the end of the day, the thing that will decide if a collaborative robot can be deployed in an application is the risk assessment.
The application needs to be proven to be collaborative through the risk assessment. The robot needs to run within the parameter sets which allow the stamp of “collaborative approval” to be put on the application itself. By not completing a risk assessment and verifying that the robot meets the parameters set forth by the standard, certain liberties are being taken, which cost dearly in the form of injury or loss of life.
Putting a robot in an application for the sake of putting a robot in is counterproductive. Collaborative, for the moment, to some means “slow and safe.” One must really consider what is the value of deploying a collaborative robot in a manufacturing environment? Saving on fencing in most cases will not make or break a return-on-investment (ROI) calculation. There needs to be other factors which make deploying the robot make sense. Interaction with human co-workers and a perceived flexibility are currently the hottest items causing companies to look into collaborative robots; however, if the application doesn’t pass the risk assessment, then it should not be collaborative.
Maximiliano Falcone is senior manager, general industries engineering at Kawasaki Robotics USA.
Patrick Laughter, Denso Products & Services Americas: While collaborative robots offer interaction with humans and added savings from not having safety guarding or expensive circuitry, they can only operate at 250 mm/sec cycle rates, and current is limited. This increases cycle time, and payloads need to be lighter.
Patrick Laughter is engineering manager, robotics, Denso Products & Services Americas.
Brian Carlisle, Precise Automation: The robot must be collaborative to enable a collaborative application. However, that is not sufficient. The application tooling must also be safe for collisions with people. A robot with a needle or a cutting tool will need work-cell guarding, even if the robot is collaborative. In general, collaborative robot applications are limited to payloads of a few kilograms. At more than 4 or 5 kg, the motion must be restricted to 500 mm/sec or less in order to limit the kinetic energy of the robot and payload, so that a collision does not injure a person. However there are many, many applications that only require a payload of a few kilograms. GM did a study of the weight distribution of the parts in a car and found that 95% weighed less than 5 lb, even though there are a number of large, heavy parts. While collaborative robots are typically less expensive to install, since they do not need elaborate safety screens and interlocks, and as a result take less floor space, they are slower than standard robots. Some robots can move at tool speeds approaching 10 m/s, whereas collaborative robots typically work at tool speeds of 500 mm/s to 1.5 m/s. Note also however, that many applications are paced by a machine or conveyor cycle time, which can be several seconds or longer, so that the slower speed of collaborative robots is not a limitation for throughput. So, as users understand collaborative robots, they are finding many applications where lower cost, faster installation, less floor space and often easier teaching far outweigh the blazing speed of conventional robots.
Brian Carlisle is CEO at Precise Automation.
Travis Schneider, Parker Hannifin: The robot is enabled to be collaborative, but in my opinion the application should be conducive to a collaboration between worker and robot for a customer to take full advantage of the robot. Collaborative robots often operate at reduced speed or payload to minimize the potential pain or injury it could do to a human. Nevertheless, the application must be scrutinized against risk assessments that act as general guides to ensure safe performance. Even when operating at reduced payloads or speeds, if the collaborative robot is wielding a knife or needle, it could still likely inflict a lot of harm to an unsuspecting passerby.
Most applications for robots fall into one of the three Ds of dull, dirty or dangerous—of these, the dull or mundane operations are probably some of the best targets for collaborative robotics. However, if a robotic operation is to be performed away from people or, more importantly, can keep the human away from a dirty or dangerous operation, then a “standard” industrial robot may be a better option as it won't have speed or payload limitations. This is especially true when one is automating a process to make large throughput gains. Another general guide to keep in mind is to use the right tool for the right job. In the same way that not every application requires a collaborative robot, not every application requires a six-axis arm. There are tons of different configurations for robotic manipulation. I often try to guide engineers to deploy the right number of degrees of freedom for what the application requires. In addition to potentially being less expensive, constraining the motion to the proper number of degrees of freedom can actually provide a higher degree of precision in motion.
An uncaged industrial robot requires an extensive amount of sensing—LIDAR, infrared—and control integration, and it is likely cost-prohibitive. However, as the cost of sensors continues to fall, this will likely be a viable collaborative method.
The only downside I could see would be throughput or payload limitations, as mentioned. This is again assuming the application requires a six-axis operation. , Going back to my earlier comment, in using the right degrees of freedom for the application, a lower-degree-of-freedom system could be more precise, smaller and less expensive than a collaborative robot.
Travis Schneider is IoT business development, automation group at Parker Hannifin.
Mike Van Hoomissen, DWFritz Automation: Collaborative robots are great for low-speed, repetitive tasks that require close contact with humans. The safety aspect of collaborative robots limits their speed and power. We typically use wafer, gantry, four-axis SCARA and six-axis robots in safety cages that can run much faster with higher payloads. One example of using collaborative robot is management of long-wire pigtails of instruments for soldering. While the soldering operation is intense on the human eye and produces harmful fumes, the pigtail handling would be easy for a human. So the collaborative robot can perform the soldering while the wire-harness management can be dealt with by the human. If the distance moved by the robot arm is short, then the slowness factor becomes less of an issue.
Mike Van Hoomissen is senior staff software engineer at DWFritz Automation.
Andie Zhang, ABB: In current ISO standards 10218-1 and TS15066, there is no definition for collaborative robots, but only for collaborative applications, so it is the application that is collaborative, and different robots, including conventional robots can be used. What’s new is that through different safety functions, such as speed and separation monitoring and smaller, lighter robots, robots and humans can work closer together than before. Through power and force limiting by inherent design or control, it’s possible for a robot to make safe contact with humans, though this contact force is depending on the payload, speed and geometry of contact surface, so it’s important to always make a risk assessment for the application. Generally, collaborative robots are designed to do what human operators do, and they will generally be slower than conventional robots and have a lower payload due to safety concerns; therefore, if the application does not require any sharing of workspace between operator and robot, it may be best to use a conventional robot. That said, collaborative robots open up enormous possibilities to automate tasks that have not been able to be automated before, and they enable companies that have never used industrial robots before to adopt them easily and affordably.
Andie Zhang is global product manager, collaborative robotics at ABB.
Daniel Moore, Universal Robots: As the industry leader in collaborative robotics, we acknowledge that it is the application that is collaborative, but you need collaborative systems to build that application. There are four types of collaborative devices, but we consider the highest and most difficult standard, permanent force- and power-limited devices, to be true “cobots.” Most industrial robots are so strong and so massive and so insensitive to outside forces that there is no chance to avoid catastrophe in the event of a high-speed collision, regardless of the type of payload. A true collaborative robot in an assessed collaborative application should be able to move at high speeds without endangering nearby humans.
Now, traditional industrial robots will always have the advantage in speed and payload capacity, but we believe one of the main selling points of Universal Robots is ease-of-use and redeployability; even if a line is shut down, that robot can be easily unmounted and moved by hand to a new spot and quickly be reprogrammed to continue producing profit for the owner.
Daniel Moore is tech support manager at Universal Robots.
Bhaskar Ramakrishnan, DWFritz Automation: In general, robots can operate at a faster speed and without fatigue, compared to human beings. Depending on the application and speed requirement, a collaborative robot may be used. Other demanding environments that are hot or corrosive, or where repetition and heavy lifting play a role and which need some sort of human intervention, can utilize a collaborative robot. Note that the cost of a collaborative robot is higher compared to a regular robot with a safety cage. The safety cage does take more space, so collaborative robots can be space-saving.
Bhaskar Ramakrishnan is technical sales engineer at DWFritz Automation.
Jim Lawton, Rethink Robotics: Collaborative robots will not displace the need for traditional industrial automation. There will always be tasks that require heavier payloads and more speed than a collaborative robot is suited to perform. In the end, the application is important, and collaborative robots are built to fill the gaps in the factory. Boston Consulting Group estimated that only about 10% of automatable tasks are actually automated in a factory. Collaborative robots are designed to automate those tasks, where working uncaged and collaboratively with factory workers is a requirement. Because collaborative robots are so easy to train, relatively inexpensive and can be moved around the floor because they don’t have cages to contain them, the robot may be used on a wide variety of applications and more easily justified. This is one of the many benefits that collaborative robots have over traditional robots.
Jim Lawton is chief product and marketing officer at Rethink Robotics.