Efficient manufacturing calls for coordinated systems of shop personnel, equipment and software. These systems increasingly include robotic technology, as manufacturers recognize the reliability, repeatability and flexibility that robots provide. According to the International Federation of Robotics, the number of industrial robots in use worldwide will increase to around 2.6 million by 2019, about one million more than in 2015. Approximately 70% of industrial robots are used in the automotive, electrical/electronics and metal and machinery industries.
Safety is the first consideration in any manufacturing operation; safe use of industrial robots traditionally has required maintaining set distances between shop-floor personnel and robots, and positioning guards and cages around the robots to prevent worker injuries. A growing assortment of robots, software and implementation strategies is aimed at closer and more productive interaction between humans and robots.
Control of this relationship ranges from sensor-governed interruption of a robot’s movement when workers move into potentially dangerous areas, to application of fully “collaborative” robots, also called cobots, engineered to operate safely side-by-side with human workers. Differing modes of collaboration enable manufacturers to take advantage of the full capabilities of robotic systems and shop personnel.
The Dull, Dangerous & Dirty
The current shortage of skilled workers in the US drives much of the demand for collaborative robots. The retirement of veteran shop personnel and not enough younger workers entering the manufacturing workforce has left shop owners with staffing shortfalls.
To address this skills gap, manufacturers are using robots for repetitive and/or dangerous operations and moving workers to jobs requiring more thought and skill. “We are seeing that people get reallocated,” said Craig Tomita, sales manager, western region, for Universal Robots USA Inc. (Ann Arbor, MI), a maker of fully collaborative robots.
Tomita said implementation of collaborative robots is relatively simple because they work in a human-scale envelope and don’t require the guarding necessary with traditional industrial robots. There is no need to rearrange the shop floor and space utilization doesn’t change. “You just put collaborative robots where human operators are currently doing the work,” he said, adding that while collaborative robots are safe, risk assessment — determining all potential hazards in a system and finding ways to mitigate them — is vital.
Ease of use is a key attribute of fully collaborative robots. Although the capabilities of small collaborative robots are on a par with those of small traditional robots, collaborative training and programming requirements are minimal. “The ease of programming in the UR system is a game changer,” said Tomita. “When, for example, a robot needs to be integrated into a larger manufacturing system, Universal’s units can be programmed and run using a hand-held touch screen. If needed, higher-level users can also program the robot using its Python-like scripting language.”
Collaborative robots can be easily used by smaller shops, where integration costs have been a barrier to entry. “If you look at it like an iceberg, the robot part is peeking up above the water,” said Tomita. “The integration process is what is below. You take that iceberg as whole and it gets really expensive. A collaborative robot costs about the same as a standard industrial robot, but the integration cost is much lower,” he said.
Utilization of robots can also help reduce labor costs, Tomita said: “Companies that use collaborative robots often can quote jobs that they weren’t able to before because their labor costs were out of the ball park.”
Potential applications are wide open. “The number and kinds of applications for our collaborative robots have yet to be thought of,” Tomita said. “People are so used to thinking that the robot is behind a cage over there, the people are over here, and the two shall never meet. That is changing.”
Universal Robots offers three models of six-axis collaborative robots, from UR3 units that handle payloads up to 3.3 kg (7.26 lb) and have a working radius of 500 mm (19.7″), to UR10 robots with a 10-kg (22-lb) payload and working radius of 1300 mm (51.2″).
A typical Universal robots customer is CleanLogix LLC (Santa Clarita, CA), which develops products that use CO2 in cleaning, cooling and other applications. CleanLogix President David Jackson said, “We can broadcast spray into a cutting zone and provide a very clean and green means for cooling a cutting tool in a process.” He noted that a single robot is able to perform different tasks; for example, placing a part in a machining center then changing end-effector tooling to provide coolant. “So you have a multitasking robot and a total green, lean solution. We are very excited about these small-form-factor robots.”
The ability of collaborative robots to work safely with humans enables operations to be organized to maximize synergy; for example, a robot can assemble a series of components while a human worker performs insertion of tiny wires that require dexterity and cognitive perception. Robots are ideal for operations that require repetition, reliability and accuracy, including packaging, palletizing, assembly and pick-and-place operations.
Repeatability and accuracy are separate but related measurements of robot performance. Repeatability is a measure of a robot’s ability to return to a given position time after time. Claimed repeatability of various collaborative robots is in the range of about ±0.02 to ±0.1 mm (0.0008 to 0.004″) or lower.
Accuracy, on the other hand, measures how close to a desired position or path a robot can move, and can be categorized as positional accuracy and path accuracy. Tasks like drilling, where the robot moves to a position and stops while the hole is drilled, require positional accuracy. Path accuracy is required for processes like laser cutting or painting where the process takes place while the robot moves between points.
Application Determines Collaboration
In general, side-by-side activity of shop personnel and robots is achieved with collaborative robots that handle small payloads at slow speeds — typically in the vicinity of 1 m/sec (39.4 ips). However, said Nicolas De Keijser, assembly and test business line manager for the Robotics and Motion Division of ABB Inc. (Cary, NC), the load a robot carries or the speed at which it travels do not by themselves guarantee safe, collaborative operation. A very light but dangerous payload might be a razor blade or a hypodermic needle. A slow-moving, but heavy payload could pose a crush hazard to shop-floor personnel. Similarly, gripper failure when moving a medium-sized payload at moderate speed could send the payload hurling across the shop.
A robot system is only collaborative relative to its application. “If the operation is dangerous, you don’t want to be near the robot, no matter how collaborative the robot is,” De Keijser said, adding that a complete solution safety assessment is required in any application. “If you read the standards, they always dictate that we are talking about collaborative operation, not a collaborative robot.”
ABB takes two approaches to assuring safe coexistence of workers and robots. Truly collaborative applications, such as small parts assembly, generally can be fulfilled by the company’s table-mounted, dual-arm YuMi robot. The 38 kg (84 lb) robot’s reach is 599 mm (24″) and payload 500 kg (1.1 lb) per arm. It occupies a human-scale area while operating safely alongside shop personnel.
On the other hand, ABB facilitates safe operation of large, standard industrial robots via redundant systems that combine careful programming and safety monitoring software. After a robot is programmed for a safe range of operation, ABB’s SafeMove2 safety-certified robot monitoring software provides redundant safety protection by monitoring robot activity, including safe speed limits, standstill monitoring and axis ranges, as well as position and orientation supervision. The software is integrated into the robot controller, and can save setup time and facilitate greater productivity while lowering the total cost of investment, according to De Keijser.
Yaskawa Motoman’s Kinetiq Teach system is an example of hand guiding mode of human/robot interaction described in ISO standards 10218-1 and 10218-2.
Built-In Adaptive Control
Another form of robotic collaboration combines the skills and experience of a craftsman with the ability of a robot to operate in dangerous areas. Welding generally is not a user-friendly operation. According to Zane Michael, director of thermal business development for Yaskawa America Inc., Motoman Robotics Division (Miamisburg, OH), “In today’s market, the word collaborative means you’ve got a robot and operator zone that are overlapped. They are working closely together. I have not seen the word collaborative in that sense being applied to the welding environment.”
Although robotic welding cannot strictly be called collaborative, a form of collaboration between a skilled welder and a robot supports consistent performance. It’s not that running a welding robot is overly difficult. “Programming and running the robot is easy,” Michael said. “At Motoman, we have a system called Kinetiq Teach, where you can grab the torch and move the robot through the path you want it to follow, record that path, and you are ready to go.” However, he added, running a welding robot without knowledge of welding is “an uphill battle” if episodic weld defects such as undercuts or burn-throughs occur.
In those cases, an experienced welder can adjust the robot’s actions to overcome the problems. Adaptive control systems using lasers or other sensors are available for basic seam tracking, but the majority of welding systems are shipped without such systems. “The human welder has what I call built-in adaptive control,” Michael said.
Depending on operational complexity and volume capabilities, robot cells possess different forms of automation and safety systems. Motoman’s ArcWorld single-station C30 cell is for low-volume, single-piece processing of smaller parts. When the cell door opens to allow an operator to load a weldment on to the positioner in the cell, the robot is in emergency stop mode. After the operator leaves the cell, the door descends and welding takes place. On the other hand, Motoman’s high-volume, large-part 6000 series machines feature a positioner that indexes part fixtures like a Ferris wheel and can be loaded from outside the cell while multiple robots weld an assembly inside.
Welding companies face labor shortages similar to those in other manufacturing segments. Based on US Bureau of Labor statistics, the American Welding Society estimates a welder shortage of 290,000 in the US by 2020. Michael said forward-thinking welding suppliers are working with trade schools and career centers to create curriculum for welders. For example, Motoman has created a welding STEM program and matching curriculum for both high school and adult welding training.
Robots, the IoT
Cloud computing and the Internet of Things facilitate collaboration between robots, users, and robot manufacturers. Manufacturers routinely monitor robot uptime to document productivity, said Mark Scherler, general manager, materials joining segment, for robot supplier FANUC America Corp. (Rochester Hills, MI), “But there is more to it than that. We are using the Internet to collect data from the robots and help manufacturers improve uptime.”
FANUC provides collaborative as well as standard industrial robots. To fully exploit the data the robots collect, the company has a developed a zero downtime (ZDT) diagnostic application that detects and analyzes critical information regarding a robot’s mechanical operation and maintenance status. For instance, ZDT installed on robots in a manufacturer’s facility can sense a robot experiencing increasing torque levels that may indicate a problem with a particular axis. Via cloud technology, the ZDT application sends the operational information to the FANUC data center for analysis. Critical problems trigger transmission of notices to designated smart devices at the robot user.
At the same time, FANUC confirms availability of parts that may be required to resolve the issue. The data can also be used to optimize robotics systems in terms of improving cycle time, reducing energy consumption, and extending robot life via better maintenance.
Jeff Estes, eastern regional sales manager for the Morris-South division of the Morris Group (Charlotte, NC) and formerly director, partners in THINC for Okuma America Corp., said the high-volume automotive industry has led adoption of automation for decades. Today, smaller volume manufacturers seek the reliability and predictability provided by automation, but also want the flexibility to enable quick changeovers from one production lot to another. To that end, robotic technology allows a shop to quickly change programming details and end effectors.
Estes pointed out that robotically automating a machine tool does not simply involve loading and unloading parts. “The key is how to achieve low- or no-attendance operation,” he said. In a fully automated cell, numerous elements, including a machine tool, robot, gage, vision system, and material handling equipment, are brought together to create a “closed-loop operation that can continue to run and make basic decisions without human interaction,” Estes said. The multiple elements of a cell generate information about individual operations but, “until they start working with each other those are just pieces of information.” An integrator chooses and arranges cell elements and coordinates their functions.
Some machine tool suppliers maintain in-house integration capabilities in order to provide turnkey systems. Methods Machine Tools Inc. (Sudbury, MA) has more than 30 automation engineers in the US, as well as design, control, electrical/mechanical, systems integration and field service/installation personnel. Complementing its activities as a machine tool provider and systems integrator, Methods recently added an automation and integration center to its Charlotte, NC facility. Automation specialists and systems integration engineers provide customer consultation regarding equipment and capabilities ranging from defining and building cells to performing run-offs. The focus of the new 10,000 ft2 (929 m2) facility “is on providing a custom-tailored, comprehensive solution for our customers,” said Methods Automation Manager John Lucier.
Technologies like collaborative robots are changing the way manufacturers think about automation. Estes encourages manufacturers to leverage the power of the data generated and collected by an automation system.
Data analysis can contribute to other aspects of production, including product quality, tool management, material flow and logistics. When implementing robotic technology, manufacturers should take “more of a holistic approach,” Estes said. “Instead of just saving a person from loading and unloading a machine, a manufacturer should ask, ‘Can I do anything more with this robot to make it more IOT- or Industry 4.0-related as well?’” Robots and automation systems have unappreciated capabilities, he said: “Even we, as OEM providers, are learning every day how much more capable we can make them.”
What Makes a Robot Collaborative?
The term “collaborative robotics” is frequently used to describe any situation where robot action is controlled to accommodate human workers. In reality, the possible relationships between workers and robots vary widely by robot and by application.
ISO standards 10218-1 and 10218-2, “Safety Standards for Applications of Industrial Robots,” were published in 2011. They list requirements for safe design, protective measures and application of industrial robots. The standards define four different modes of safety-related interaction between shop personnel and robots.
One mode is the ability to perform a safety monitored stop. Sensors monitor a predetermined area around a robot, and robot movement stops when a human enters that area. This level of safety monitoring is satisfactory when human intervention is minimal, although frequent interruptions can decrease productivity. A second mode involves hand guiding, in which an operator teaches a robot to follow a desired path by grasping the robot arm and moving it to desired points on the path while recording the points with a teaching pendant. A force torque sensor in the robot recognizes the operator’s manual guidance. The sensor does not act as a safety system and, outside of the teaching mode, the robot must have other devices or systems in place to assure safe interaction with humans.
In the speed and separation monitoring mode, a vision system detects humans in a designated safety zone around the robot. As separation between the robot and the human decreases, the robot gradually slows; at a certain distance, robot movement stops. Then, as the human exits the safety zone, the robot resumes working at normal speed. This graduated safety mode helps maintain productivity even with worker interventions.
The fourth collaboration mode enables side-by-side activity of robots and humans and eliminates the need for protective guards and cages. The robots described are called power and force limiting robots. When sensors detect abnormal forces on the robot, such as those generated by contacting a worker, the robot immediately slows, stops or reverses. These robots described are called power and force limiting robots. When sensors detect abnormal forces on the robot, such as those operated by contacting a worker, the robot immediately slows, stops or reverses. These robots generally are smaller and less powerful than traditional industrial robots, and also have rounded shapes and enclosed joints to avoid pinching injuries.
In 2016, ISO issued technical specification ISO/TS 15066 that adds detail and clarification to ISO Standards 10281-1 & 2. TS 15066 defines collaborative robots as those designed for direct interaction with a human in a collaborative workspace where the human and robot can perform tasks simultaneously. The specification details the amount of force a robot can inflict on a human and cause no pain or injury. The force levels were determined via impact tests on specific areas of the human body. Still, it is critical to perform risk assessments of the operation being performed and the parts and tools involved in a specific application to assure absolute safety.