Skip to main content

Cobot Power and Force Limiting Under ISO 10218-2:2025 and ISO/TS 15066

In power and force limiting (PFL), contact between a robot application and a worker is expected and permitted, as OSHA describes it, when the forces and pressures of contact are limited such that there will be no injury, and the R15.06 subcommittee chair, quoted by A3, names it as one of three recognized collaborative technologies. UTEC Industrial designs, engineers, machines, fabricates, and installs custom material handling systems for aerospace and heavy industry from its Spokane Valley, WA facility, integrating Allen-Bradley PLC and motion control with in-house CNC machining, heat treating, and stress relief. This article explains what PFL is, how transient and quasi-static contact differ, where the published body-region limits come from, how a PFL application is tested, and how speed and separation monitoring compares; ISO/TS 15066:2016 is still current and is to be revised, and the article treats it that way. A heavy-handling cell is built along one chain, design → engineering → parts machining → fabrication → assembly → weld fatigue → stress relief → drives → controls → tuning → monitoring, and a collaborative application depends on the gripper and fixtures built upstream as much as on the robot's controls.

What does "collaborative application" mean in the 2025 standards?​

The 2025 parts of ISO 10218 drop the robot-level term. A3's FAQ says the terms "collaborative robot" and "collaborative operation" "will not be found in the revised ISO 10218. 'Collaborative application' is used instead, as only the actual use of the robot can be designed, tested, and confirmed as a collaborative application." The R15.06 subcommittee vice chair, Bill Edwards, is quoted in A3's Industry Insights: "It is the entire Collaborative application that must be safe."

The word "cobot" persists in the market. OSHA's Technical Manual chapter notes that "Some robots are referred to as 'cobots' in an effort to state that the robot is 'ready' or 'enabled' to be used in a collaborative application," and defines collaborative industrial robot applications as "those that are designed for direct interaction with workers." NIOSH's robotics overview uses a wider grouping, in which collaborative robots include wearable robotics or powered exoskeletons and co-existing or mobile robots; that is NIOSH's grouping, not the ISO 10218 term.

As engineering reasoning, the shift matters for a purchase: a robot sold as collaborative-ready becomes part of a collaborative application only after the gripper, the part, the fixture and the task have been assessed together. The 2025 definition of a collaborative application is set in Part 2's terms clause (A3 Online Marketing Team 2025, FAQ 6; Heater 2025, A3 Industry Insights, "Word Matters"; OSHA Technical Manual, Sec. IV Ch. 4, 2026, §IV and note 4; NIOSH 2024, Types of robots; ISO 10218-2:2025).

Which collaborative technologies do the 2025 standards recognize?​

The R15.06 subcommittee chair, Todd Dickey, is quoted in A3's Industry Insights: "The recognized collaborative technologies are hand-guiding controls, speed and separation monitoring, and power and force limiting." A3's 2026 summary adds that "these approaches do not eliminate the need for risk assessment." ISO 10218-2:2025 sets out these three technologies in its collaborative-application clauses.

OSHA's chapter, written against the 2012 edition, describes each:

  • Speed and separation monitoring (SSM). A protective device detects the intrusion of workers; "At a minimum, the robot application stops during the intrusion," and some integrations slow down on first detection and stop before contact. OSHA adds that "when speed is being used for safety purposes, the speed should have an associated safety function that monitors that the needed speed will not be exceeded."
  • Hand-guided controls (HGC). The robot system moves under a worker's direct control in automatic mode. OSHA's example is a worker guiding a robot to grasp a heavy box and place it onto a truck; the robot "is doing all the heavy lifting, but will not move without the worker physically directing it while the worker presses (or actuates) a hold-to-run control device."
  • Power and force limited (PFL). Contact between the robot application and a worker "is expected and permitted," but only "when the forces and pressures of contact are limited such that there will be no injury to the worker(s)."

OSHA also notes that ANSI/RIA R15.06-2012 referred to the safety-rated monitored stop as a fourth collaborative type, used only in conjunction with SSM, HGC or PFL; A3 says the 2025 term is "monitored standstill" (Heater 2025, A3 Industry Insights; A3 Online Marketing Team 2026; OSHA Technical Manual, Sec. IV Ch. 4, 2026, §VI; ISO 10218-2:2025).

Is ISO/TS 15066:2016 still current after ISO 10218-2:2025?​

Yes, according to iso.org. The ISO/TS 15066:2016 page says the publication "was last reviewed and confirmed in 2022. Therefore this version remains current," and it shows the document at stage 90.92, "To be revised." It says the specification "Will be replaced by" ISO/AWI 15066-1, Robotics — Physical contact with robots — Part 1: Biomechanical thresholds and data, a project still under development. The 2016 abstract says the specification supplements the collaborative requirements and guidance in ISO 10218-1 and ISO 10218-2, applies to industrial robot systems as described in those standards, and does not apply to collaborative applications designed prior to its publication.

A3's FAQ says that safety requirements for collaborative applications, "formerly, the content of ISO/TS 15066", have been incorporated into ISO 10218-2:2025, and the iso.org FAQ for Part 2 says the standard "complements standards such as ISO/TS 15066." In the United States, A3's store page describes RIA TR R15.606-2016 as "a U.S. National Adoption of ISO/TS 15066." Which body-region values and test provisions Part 2 itself now holds, and which it leaves to the technical specification, is set out in Part 2's PFL clauses and annexes (ISO/TS 15066:2016; ISO/AWI 15066-1; A3 Online Marketing Team 2025, FAQ 5; ISO 10218-2:2025; RIA TR R15.606-2016).

How does power and force limiting work?​

OSHA's chapter says PFL capability can be provided in two ways: "One is by inherently safe design of the robot (e.g., low energy potential due to very low payload and/or speed capability). Another is by control means," using safety functions with sensors and safety-related parts of the control system, for example "torque sensors on all joints to safety logic that will slow or stop the robot." It names "speed limiting, force limiting, and power limiting" as the typical safety functions, and says "Collaborative applications using PFL robots usually operate at much lower speeds and payloads than they are physically capable [of]," so that on contact "the robot is not moving with enough energy to cause injury."

In the specification as Behrens et al. describe it, the criterion is the onset of pain, not injury. The authors, a team from Fraunhofer IFF, Otto von Guericke University Magdeburg and BGHM, describe ISO/TS 15066:2016 this way: "any unintended human-robot contact needs [to] be considered as a hazard, which must not cause biomechanical stress beyond the onset of pain. Injuries, even slight ones, are not allowed at all." OSHA adds that robot contact with sensitive body regions, such as "the face, temples, and throat", is to be prevented or avoided per RIA TR R15.606-2016.

A3's 2026 summary describes the process: "potential contact events are identified during the risk assessment and risk reduction is used to keep those contacts below applicable threshold values." ISO 10218-2:2025 sets the PFL requirements for quasi-static and transient contact in its own PFL clause (OSHA Technical Manual, Sec. IV Ch. 4, 2026, §VI; Behrens et al. 2022, §1; A3 Online Marketing Team 2026; ISO 10218-2:2025).

What is the difference between transient and quasi-static contact?​

OSHA's chapter defines the two contact events:

  • Transient contact "occurs when the worker's movement is not restricted at the time of contact (e.g., the worker's body part can move in free-space at the time of contact)."
  • Quasi-static contact "occurs when a worker's body part is unable to move at the time of contact due to being restricted by a fixed object (e.g., trapped or pinched between the robot and a fixture)."

Behrens et al. describe a related split, by load type (impact or pinching), as ISO/TS 15066:2016 uses it. In an impact, "the contact force applied by the robot builds up quickly and decreases again quickly after reaching its maximum." In a pinching contact the force builds up slowly, and "remains at a constant value when the robot has stopped." The specification also distinguishes contact type, semi-sharp or blunt, and per the authors, "sharp contact shapes are not allowed on the surface of a cobot."

As engineering reasoning, the failure mode a heavy-handling layout has to design out is the quasi-static one: a robot stopping against a person who is pinned between the part and a fixture, a conveyor frame or a column, where the force stays on after the robot stops (OSHA Technical Manual, Sec. IV Ch. 4, 2026, §VIII; Behrens et al. 2022, §2).

Where do the body-region pressure and force limits come from?​

Behrens et al. summarize the 2016 specification's limits and their basis:

  • ISO/TS 15066 "provides a list of biomechanical limits for 29 different body locations."
  • "One part of the limits stems from a study with 100 human subjects (Melia et al., 2019)," and "Given the design of the study, the pressure-based limits apply only to contact situations in which a semi-sharp piece of the robot clamps the human body."
  • The limits for impacts and pinching over blunt surfaces "were estimated from literature data" and "are considered preliminary and will be replaced once more reliable values become available."
  • Their abstract concludes that "Most of the limits in ISO/TS 15066 are preliminary, since they are based on unverified data from a literature survey."

Melia et al.'s abstract describes the underlying pain study, not a robot study: pressure pain thresholds of 100 healthy subjects (57 men, age 18–66 years) were assessed at 29 sites across the body with a force-controlled algometer fitted with a pressure-indicating film; 40 subjects were manual labourers; "A high variability of PPTs between subjects and different measurement sites was observed," and men and manual labourers had comparatively higher adjusted thresholds. Behrens et al. say their own choice of the 75th percentile "traces back to the limits for semi-sharp pinching of ISO/TS 15066, which are associated to the same percentile."

The specification's own table values and its treatment of transient contact are read from ISO/TS 15066:2016, Annex A (Behrens et al. 2022, Abstract, §1 and §3.3; Melia et al. 2019, Abstract; ISO/TS 15066:2016, Annex A).

What limits did Behrens et al. propose, and under what conditions?​

Behrens et al. ran a human-subject study with 112 subjects, in four regular test groups and one control group, in which "the contact force was gradually increased until the load evoked a slightly painful feeling." Their Table 8 is titled "Proposed limits for safe pHRI; pressure and force limits reflect the pain threshold of the 75th percentile of a group in which males make up 70%." Each row gives a pressure limit for semi-sharp contact and a force limit for blunt contact, for pinching and for impact. Three rows show the range:

Body location (Table 8 no.)Pinching, semi-sharp (N/cm²)Pinching, blunt (N)Impact, semi-sharp (N/cm²)Impact, blunt (N)
(2) Temple50607090
(14) Radial bone70100170180
(17/18) Forefinger pad70150240390

These are the authors' proposed values, not the values of ISO/TS 15066 or ISO 10218-2:2025, and they carry the authors' conditions:

  • Body locations. "The validity of the limits is tied to the body locations we have tested in the study."
  • One experimenter. To reduce inter-rater errors, "all tests were executed by one single experimenter," a decision the authors say "must be seen as critical" because it made that type of error impossible to trace.
  • Gender. "The pain thresholds of males and females are different in specific body regions," and the gender difference must be taken into account when limits are defined.
  • Count. The abstract reports tests at 28 body locations; Table 8 covers the body locations numbered 1 to 29, with the neck values marked as an estimate based on Melia et al.'s data, because medical concerns precluded testing the neck.

As engineering reasoning, a value from this table is a research proposal for a defined population, and a cell's acceptance limits come from the standard and the technical report it is assessed against (Behrens et al. 2022, Abstract, §3.3 Table 8, §4.1, §4.4 and §5).

How is a PFL application tested before acceptance?​

OSHA's chapter sets out the sequence. During the risk assessment, determine the expected contact areas on the worker's body, the possible contact types, and the allowable force from RIA TR R15.606-2016 for that body part and contact type. "Measure the pressure and/or other forces for each contact event prior to factory acceptance testing. Compare the measured pressure/force with the allowable pressure/force."

The chapter adds three duties:

  • Integrator. It "should use the robot manufacturer's information (i.e., moving mass of the manipulator, speed capabilities) combined with the payload mass of the end-effector and/or workpiece to determine the maximum allowable speed for contact events," and should confirm that contact events do not exceed the limits.
  • User. It "should periodically verify that the safety-function settings are still valid (e.g., speed limit has not increased or workpiece attributes changed)"; the measurements are performed by a competent worker and documented.
  • Exceedance. If a limit is exceeded, safety function settings should be adjusted by adding risk reduction measures and/or by modifying or replacing the application.

For the measurement itself, A3's store page says RIA TR R15.806-2018 "describes test methods and metrics for measuring the pressures and forces associated with quasi-static and transient contact events of collaborative applications" and gives guidance on test conditions, measurement devices and testing methods. UTEC Industrial performs factory acceptance testing (FAT) and on-site commissioning (OSHA Technical Manual, Sec. IV Ch. 4, 2026, §VIII; RIA TR R15.806-2018, A3 store page; RIA TR R15.606-2016).

How does speed and separation monitoring differ from PFL?​

In OSHA's descriptions, SSM slows or stops the robot before contact, while PFL permits contact within limits. OSHA notes that "It is common to see SSM combined with PFL for collaborative applications so that the application can run at high speed when no workers are nearby, but then slow such that contacts would be permissible according to PFL."

Marvel and Norcross at NIST describe the minimum protective distance as ISO/TS 15066:2016 presents it: the sum of the operator's travel over the robot's response and stopping times, the robot's travel over its response time, the robot's stopping travel, and margins for intrusion distance and for robot and sensor position uncertainty. They note that "T_S is a function of the robot's speed and load" and that the equation "is derived from" the ISO 13855:2010 safeguard-distance equation. Citing the 2011 edition of ISO 10218-1, they say braking time and distance result from measurements per its Annex B and "are reported as functions of the robot's initial speed, v_R, and payload (a percentage of the maximum rated load)," and they recommend that both values "be periodically reevaluated to accurately accommodate the degradation of brake components." From their earlier study they report that "even a relatively simple SSM algorithm may improve the robot system's safety, but will not provide safety in all circumstances."

The editions matter. The paper used ISO 13855:2010 and ISO 10218-1:2011; the current editions are ISO 13855:2024, which does not consider approaches such as running, jumping or falling, and ISO 10218-1:2025. The 2025 SSM formula is read from ISO 10218-2:2025, and the stopping-data annex from ISO 10218-1:2025. The article on what FANUC DCS can and can't replace covers safety-rated speed and zone checks on a FANUC controller (OSHA Technical Manual, Sec. IV Ch. 4, 2026, §VI; Marvel and Norcross 2017, §I and §IV; ISO 13855:2024; ISO 10218-1:2025; ISO 10218-2:2025).

Can power and force limiting work for a heavy payload?​

No source read for this article addresses PFL at heavy payloads; the PFL studies read for it concern cobot-scale contact. What the sources do say bounds the answer:

  • Energy. OSHA describes inherently safe PFL as "low energy potential due to very low payload and/or speed capability."
  • Speed from mass. OSHA says the integrator determines the maximum allowable speed for contact events from the moving mass of the manipulator combined with the payload mass of the end-effector and workpiece.
  • Scope of the NIST guidance. NIST's best-practice report on cobot integration is written for small- and medium-sized manufacturers.

The rest of this answer is engineering reasoning. A heavy casting, coil, log or airframe component adds moving mass the PFL limits were not built around, and the allowable contact speed falls as that mass rises; for a heavy part, PFL alone may leave too little speed to be useful. Hand-guided control is the collaborative technology OSHA illustrates with a heavy load, and SSM or conventional guarding is the alternative for automatic motion at speed. The payload, moment and inertia checks that size any robot for a heavy part are covered in the article on sizing an industrial robot (OSHA Technical Manual, Sec. IV Ch. 4, 2026, §VI and §VIII; Horst, Marvel and Messina 2021, NIST AMS 100-41).

What controls and sensing does a collaborative application need?​

OSHA's chapter says the safety functions required depend on the potential contact situations: for an application that expects contact while the robot moves, the robot selected should have PFL capabilities, while one intended to have no contact while the robot moves can use SSM or SSM/PFL. It lists other safety functions that could be required:

  • Protective stop
  • Force limiting
  • Speed limiting
  • Soft axis-limiting
  • Space limiting
  • Position limiting

It adds that "most of the safety functions required for SSM are different than those for PFL," and that an application using both needs all of them; that the manufacturer's manual "Typically" lists the safety functions provided; and that "a third-party can certify the safety functions." A3's FAQ says safety functions that enable a collaborative task "can be part of the robot or can be provided by a protective device, or a combination." OSHA's passive measures include rounded corners and edges on the end-effector and fixture, padding, compliant elements and smooth covers.

ISO 13849-1:2023 specifies a methodology and provides related requirements, recommendations and guidance for the design and integration of safety-related parts of control systems (SRP/CS) that perform safety functions, for high demand and continuous modes of operation; it does not apply to low demand mode of operation. UTEC Industrial, a Rockwell Automation Recognized System Integrator, builds the Allen-Bradley ControlLogix and CompactLogix controls that carry the cell's safety inputs, and the article on connecting a FANUC robot to an Allen-Bradley PLC covers the safety network (OSHA Technical Manual, Sec. IV Ch. 4, 2026, §VIII; A3 Online Marketing Team 2025, FAQ 4; ISO 13849-1:2023).

Where does a collaborative application sit in the build chain?​

Each link sets something the contact assessment depends on:

  • Design. The risk assessment identifies the contact events, the body regions and the contact types.
  • Engineering. End-effector and fixture geometry decide whether a contact is blunt or semi-sharp, and whether a person can be pinned.
  • Parts machining, fabrication and stress relief. As engineering reasoning, the radii and surfaces the assessment assumed have to be the ones machined and fabricated, and a fixture that distorts after welding can change a clearance into a pinch point. The article on stress relief for machine frames and bases covers why welded frames are stress-relieved before final machining.
  • Drives. Marvel and Norcross recommend periodic reevaluation of stopping time and distance as brake components degrade.
  • Controls and tuning. Speed, force and power limits are set from the measured contact values.
  • Monitoring. OSHA says the user should periodically verify that safety-function settings are still valid.

UTEC Industrial integrates FANUC robotic cells, including vision, with a FANUC design and engineering partner (OSHA Technical Manual, Sec. IV Ch. 4, 2026, §VIII; Marvel and Norcross 2017, §IV).

Related Articles

References​

  • ISO 10218-2:2025: Robotics — Safety requirements — Part 2: Industrial robot applications and robot cells. ISO, 2025.
  • ISO 10218-1:2025: Robotics — Safety requirements — Part 1: Industrial robots. ISO, 2025.
  • ISO/TS 15066:2016: Robots and robotic devices — Collaborative robots. International Organization for Standardization, 2016.
  • ISO/AWI 15066-1: Robotics — Physical contact with robots — Part 1: Biomechanical thresholds and data (approved work item, under development). International Organization for Standardization, 2025.
  • RIA TR R15.606-2016: Collaborative Robots. Robotic Industries Association/A3, 2016.
  • RIA TR R15.806-2018: A Guide to Testing Pressure and Force in Collaborative Robot Applications. Robotic Industries Association/A3, 2018.
  • A3 Online Marketing Team. Updated ISO 10218: Answers to Frequently Asked Questions (FAQs) (Robotics Blog). Association for Advancing Automation, March 20, 2025.
  • A3 Online Marketing Team. 2026 Robot Safety Standards Update: What Manufacturers and Integrators Need to Know (Robotics Blog). Association for Advancing Automation, September 4, 2026.
  • Heater, B. ANSI, A3 Publish Revised R15.06 Industrial Robot Safety Standard (Industry Insights). Association for Advancing Automation, September 10, 2025.
  • OSHA. OSHA Technical Manual (OTM), Section IV: Chapter 4 - Industrial Robot Systems and Industrial Robot System Safety. U.S. Department of Labor (undated web documentation, accessed September 2026).
  • NIOSH. Robotics in the Workplace: An Overview. National Institute for Occupational Safety and Health, CDC, February 9, 2024.
  • Behrens, R., Pliske, G., Umbreit, M., Piatek, S., Walcher, F., & Elkmann, N. (2022). "A Statistical Model to Determine Biomechanical Limits for Physically Safe Interactions With Collaborative Robots." Frontiers in Robotics and AI, 8, 667818.
  • Melia, M., Geissler, B., König, J., Ottersbach, H. J., Umbreit, M., Letzel, S., & Muttray, A. (2019). "Pressure pain thresholds: Subject factors and the meaning of peak pressures." European Journal of Pain, 23(1), 167-182.
  • Marvel, J. A., & Norcross, R. (2017). "Implementing Speed and Separation Monitoring in Collaborative Robot Workcells." Robotics and Computer-Integrated Manufacturing, 44, 144-155.
  • ISO 13855:2024: Safety of machinery — Positioning of safeguards with respect to the approach of the human body. ISO, 2024.
  • Horst J, Marvel J, Messina E. Best Practices for the Integration of Collaborative Robots into Workcells Within Small and Medium-Sized Manufacturing Operations, NIST AMS 100-41. National Institute of Standards and Technology, 2021.
  • ISO 13849-1:2023: Safety of machinery — Safety-related parts of control systems — Part 1: General principles for design. International Organization for Standardization, 2023.

Ready to Discuss a Material Handling System?​

UTEC Industrial designs, engineers, machines, fabricates, and installs custom material handling systems for heavy industry, from the stress-relieved structure and drives to the Allen-Bradley PLC controls, tuning, and monitoring that run them, at its Spokane Valley, WA facility. Send UTEC the application, loads, and duty cycle to start a system review.

Request a Quote →

Questions? Call (509) 922-1832 or email sales@utec.co

Last updated on