What are the Important Considerations when Assessing Cobot Safety?

Update: April 19, 2024

Collaborative robots (cobots) are designed to work with humans and support flexible production in Industry 4.0 factories. Compared with traditional industrial robots, cobots are simpler, easier to set up, and don’t require safely isolated workspaces. Because they are designed to work with people, cobots are built differently than other industrial robots, including features like collision detection systems, force feedback, elastic actuators, and low-inertia servo motors.

Since they are different by design, specific safety standards have been developed for cobots. The International Organization for Standardization Technical Specification (ISO/TS) 15066 specifies safety requirements for industrial cobots and their work environments. It supplements the requirements and guidance on cobot operation in ISO 10218‑1 and ISO 10218‑2.

This article briefly reviews the requirements of ISO/TS 15066 and how they fit in with ISO 10218-1 and 10218-2. It then considers the complexities of collaboration, including how the collaborative workspace is defined. It examines factors related to robot safety, like safety features built into cobots, and what external safety functions are needed, along with exemplary devices like proximity sensors, light curtains, and safety contact mats. It closes with a brief review of a few applications specific to cobot safety considerations.

There are several key safety standards for industrial robots and cobots. ISO/TS 15066 details safety requirements for industrial cobot systems and the work environment and was written to build on and supplement the limited requirements in previous standards like the ISO 10218 series. ISO 10218-1 focuses on general robots and robotic devices, while ISO 10218-2 focuses on robot systems and integration. American National Standards Institute/Robotics Industry Association (ANSI/RIA) R15.06 is a national adoption of ISO 10218-1 and ISO 10218-2.

Complexities of collaboration

Before getting into the details of cobot safety, it’s helpful to define collaboration. Collaboration in robotics is complex and includes three factors:

  • A cobot is a “robot designed for direct interaction with a human within a defined collaborative workspace,” according to ANSI/RIA R15.06.
  • A collaborative operation is a “state in which a purposely designed robot system and an operator work within a collaborative workspace,” according to ISO/TS 15066.
  • Finally, a collaborative workspace is the “workspace within the safeguarded space where the robot and a human can perform tasks simultaneously during production operation,” according to ANSI/RIA R15.06.

It comes down to the definition of the collaborative workspace “within the safeguarded space.” The safeguarded space includes a layer of safety protection in addition to the standard safety functions included in the cobot.

Common protective features integrated into cobots include contact detection systems based on torque measurements at every joint that monitor for unexpected impacts, obstructions, or excessive forces or torque. There should also be automatic braking systems and manual brake releases for moving the arm without power.

Unexpected contact with the person by the cobot is a particular concern. The standards dictate that contact should be prevented anywhere on a person’s head. In addition, the standard splits the body into 29 specific areas and details limitations for two types of contact:

  • Transient contact is a moving, dynamic event where the cobot hits a person. Limitations are based on location, inertia, and relative speed.
  • Quasi-static contact occurs when a body part is trapped between the cobot and a surface. Limitations are based on pressure and force related to crushing and clamping effects.

The specification provides guidance, not absolute limits, based on application considerations. It also states that the guidance is informative and reflects current best practices since collaboration between people and robots is a new field, and research is ongoing.

Continuum of collaboration

There is no single collaborative application. People and cobots can interact and collaborate in a continuum of ways. Collaborative applications range from co-existence, where a robot stops under power when a person enters the collaborative workspace, to an interactive activity with the person touching the cobot while in operation (Figure 1).

Figure 1: Human and robot collaboration includes a broad range of possible levels of interaction. (Image source: SICK)

A risk assessment is required to identify the safety needs of individual collaborative applications. It includes identifying, evaluating, and reducing the hazards and risks associated with the application. ISO 10218 includes a list of safety features that can be appropriate in various circumstances but no definitive requirements. ISO/TS 15066 brings additional details to cobot risk assessments. In each case, the goal of the risk assessment is to identify external safety devices and systems needed to ensure the safe implementation of collaborative applications.

For a deeper dive into risk assessment and robots, see the article “Safely and Efficiently Integrating AMRs into Industry 4.0 Operations for Maximum Benefit.”

Protection and efficiency

While cobots are designed for safe operation, additional protection layers can improve collaborative applications’ efficiency. Without additional safety, when a person enters the collaborative workspace, ISO/TS 15066 mandates a maximum speed of 0.25 meters per second (m/s) per axis. For most cobots, that’s very slow.

For example, the LXMRL12S0000 Lexium cobot from Schneider Electric has a maximum payload of 12 kilograms (kg), an operating radius (working range) of 1327 millimeters (mm), positioning accuracy of ±0.03 mm, and a maximum speed of the tool end of 3 meters per second (m/s), 12 times faster than the maximum allowed by ISO/TS 15066 when a person is in the collaborative workspace (Figure 2).

Figure 2: This cobot can move 12 times faster than the maximum allowed by ISO/TS 15066 when a person is in the collaborative workspace. (Image source: Schneider Electric)

In many applications, the cobot can be operating alone for long periods. So, sensing the presence or absence of people in the collaborative workspace can enable much faster operation and higher efficiency when no one is present. Common devices for sensing the presence of people include safety scanners, light curtains, and safety contact floor mats. Each technology offers a different set of benefits, and they are often used in combination.

Safety scanners

Safety scanners monitor a designated area to detect the presence of people. They can determine how far away a person is and implement various warning zones in addition to the active safety zone.

Omron’s model OS32C-SP1-4M is a good example of a safety laser scanner designed for use with cobots. It has a safety radius of up to 4 meters (m) and can support multiple warning zones up to 15 m. It includes 70 standard sets of safety zone and warning zone combinations to support complicated collaborative workspaces. In addition, the minimum object resolution can be set to 30, 40, 50, or 70 mm, and the response time can range from 80 milliseconds (ms) up to 680 ms, further increasing application flexibility (Figure 3).

Figure 3: This safety scanner has a safety radius of up to 4 m and can support multiple warning zones up to 15 m. (Image source: DigiKey)

Light curtains

Light curtains can measure the presence of people and can be designed to detect objects of various sizes, like fingers or hands. Unlike safety scanners, light curtains don’t measure distance. They send a series of light beams between linear emitter and receiver arrays and can sense when an object breaks one or more beams.

In terms of safety ratings, there are two primary light curtain classifications: Type 2 and Type 4. They have similar outward appearances but are designed to provide different levels of safety. Type 4 monitors the safeguarded space that defines a collaborative workspace. Type 2 light curtains are designed for lower-risk applications.

Light curtains guard perimeters and are available with several levels of resolution, like 14 millimeters (mm) for finger detection and 24 mm for hand detection. The model, SLC4P24-160P44 from Banner Engineering, is a Type 4 light curtain kit with an emitter and receiver array and has a resolution of 24 mm to protect people and machines like cobots (Figure 4). The emitters have a row of synchronized modulated infrared light-emitting diodes. Receivers have a corresponding row of synchronized photodetectors. The emitters have a 2-meter range, and these light curtains can be installed in lengths from 160 to 320 mm in 80 mm increments.

Figure 4: This Type 4 light curtain has a resolution of 24 mm. (Image source: Banner Engineering)

Safety laser scanners and light curtains provide non-contact means for enhancing the safety of collaborative workspaces. However, they can be difficult to use in optically challenging environments like areas with highly reflective surfaces that can send unwanted light interference, and they can trip because of leaking oil or grease or excessive dust or humidity.

Some of these optical sensors include sensitivity adjustments that can help mitigate certain types of interference. Those sensitivity adjustments can also increase response times and other performance compromises. Another solution is to use a safety contact mat together with optical sensing devices.

Safety contact mats

Safety contact mats have two conductive plates separated by a rasterized insulating layer and can be used alone or in combination with other types of sensors. If a person steps on the mat, the top conductive plate is depressed and contacts the lower plate, triggering an alert signal (Figure 5). The exterior of the mats is a polyurethane material that’s slip-resistant and impervious to water, dirt, and oil. The SENTIR mat model 1602-5533 from ASO Safety Solutions can connect up to 10 mats in series to a single monitoring unit for a maximum coverage of 10 m2.

Figure 5: When stepped on, the safety mat’s top and bottom conductive layers make contact, triggering an alert signal. (Image source: ASO Safety Solutions)

Safety is in the details

There is no single formula that guarantees safety. Every collaborative application is different and needs to be handled based on its unique characteristics and needs. A key factor is: where does the application lie on the continuum of collaboration (see Figure 1)? The closer the interaction between the cobot and people, the more safeguarding is needed.

There are more details to consider. Some of them include:

  • Each location needs to undergo a detailed risk assessment to see if the cobot has been moved from workstation to workstation. Even if they appear to be the same, small variations can make a difference in safety.
  • If other machines are in the collaborative workspace, do they need to be linked to the shutdown system or the safety slowdown for the cobot?
  • This article has focused on safety-related hardware, but for networked systems that are increasingly common, cybersecurity is an important consideration to prevent interference with cobot operation or the safety systems.

Conclusion

Cobot safety is complex. It begins with defining the collaborative workspace within the safeguarded space and requires a risk assessment of the collaborative operation. Standards like ISO/TS 15066 and the ISO 10218 series are important and provide recommendations and guidelines. Cobots include basic safety features like collision detection systems, force feedback, elastic actuators, and low-inertia servo motors. Depending on the specifics of the collaborative application, additional safety devices like proximity sensors, light curtains, and safety contact mats may be needed.