Human-Robot Collaboration Safety Assessment: Standards, Testing Items, and Compliance Methods

Human-Robot Collaboration (HRC) safety assessment is a systematic evaluation process that verifies whether collaborative robot systems can operate safely alongside human operators in shared workspaces. Guided by ISO 10218, ISO/TS 15066, and related functional safety standards, it covers risk assessment, power and force limiting validation, speed and separation monitoring, and safety control system integrity. This article presents a complete technical overview of HRC safety assessment, including applicable standards, testing samples, testing items, testing methods, and typical industrial applications.

Technical Overview

Human-Robot Collaboration (HRC) safety assessment is the structured process of evaluating, validating, and documenting the safety performance of robotic systems in which humans and robots share a common workspace without conventional physical guarding. Unlike traditional industrial robot cells that rely on fences and interlocked gates, collaborative applications must guarantee safety through the design of the robot itself, its control system, and its operational parameters. The assessment determines whether these engineered protective measures reduce risk to an acceptable level during all phases of operation, including normal production, setup, maintenance, and foreseeable misuse.

The technical framework for HRC safety assessment is built on a hierarchy of international standards. The revised ISO 10218-1 and ISO 10218-2 (2025 editions) define safety requirements for industrial robots and robot systems respectively, and have integrated the collaborative operation requirements originally published in ISO/TS 15066. Supporting standards address risk assessment and functional safety, forming a complete compliance chain for manufacturers, integrators, and end users.

• ISO 10218-1 / ISO 10218-2 — Safety requirements for industrial robots and robot systems, including collaborative operation
• ISO/TS 15066 — Collaborative robot technical specification, including biomechanical force and pressure limits
• ISO 12100 — Safety of machinery: risk assessment and risk reduction methodology
• ISO 13849-1 / IEC 62061 — Functional safety of safety-related control systems (Performance Level PL, Safety Integrity Level SIL)
• ISO 13855 / ISO 13857 — Positioning of safeguards, safety distances, and minimum gaps
• ISO 13850 — Emergency stop function principles for design
• ANSI/RIA R15.06 and the EU Machinery Regulation — Regional market access and regulatory compliance

ISO 10218 and ISO/TS 15066 define four collaborative operation techniques, each requiring distinct verification strategies: safety-rated monitored stop, hand guiding, speed and separation monitoring, and power and force limiting. A real-world application may combine several of these techniques, and the safety assessment must validate each mode as well as the transitions between them.

Testing Samples

Collaborative Robot Arms and Controllers

Typical samples include six-axis and seven-axis collaborative robot arms, dual-arm cobots, and their associated safety controllers. Testing verifies joint torque limiting, compliant behavior, collision detection sensitivity, and the integrity of the safety-rated control architecture, which is typically required to achieve Performance Level d with Category 3 architecture under ISO 13849-1.

End-Effectors and Collaborative Tooling

Grippers, vacuum end-effectors, pneumatic screwdrivers, sanding and polishing spindles, and sensor-guided tools are evaluated as part of the complete collaborative application. Because the end-effector defines the actual contact geometry during a collision, its mass, sharpness, stiffness, and surface characteristics directly influence the achievable force and pressure levels.

Integrated Collaborative Workcells

Fully assembled human-robot workstations, such as machine-tending cells, assembly stations, and kitting systems, are assessed at the system level. System-level evaluation is mandatory because collaboration safety depends on the combined behavior of the robot, tooling, workpiece, fixtures, and peripheral safety devices rather than on any single component.

Safety-Rated Peripherals and Interaction Devices

Safety laser scanners, light curtains, safety mats, 3D camera systems, safety PLCs, force-torque sensors, and hand-guiding devices are tested for detection capability, response time, and interface integrity. These devices implement the speed and separation monitoring and safety-rated monitored stop functions required in shared workspaces.

Mobile Manipulators

Collaborative robot arms mounted on autonomous mobile robots (AMRs) are evaluated for combined hazards, including navigation safety, braking performance, load stability, and dynamic protective separation distance calculations that account for the motion of both the base and the arm.

Testing Items

A comprehensive HRC safety assessment addresses the following core evaluation items:

• Risk assessment and risk reduction verification according to ISO 12100, covering all operating modes, tasks, and hazard zones
• Power and force limiting validation, including both quasi-static (clamping or squeezing) contact and transient (dynamic impact) contact
• Force and pressure measurement against the biomechanical limit values defined for specific human body regions in ISO/TS 15066 Annex A
• Speed and separation monitoring performance, including protective separation distance calculation and dynamic robot speed reduction
• Safety-rated monitored stop verification, confirming complete standstill before human entry into the collaborative workspace
• Hand guiding function safety, including guiding device force limits and emergency stop accessibility
• Robot stopping time and stopping distance measurement at rated load and maximum speed conditions
• Functional safety verification of the safety control system, including Performance Level (PL) or SIL verification, fault reaction behavior, and diagnostic coverage
• Validation of limiting functions: axis space, Cartesian space, and orientational workspace limits, as well as safety-rated speed monitoring
• Collision detection threshold setting verification and response time measurement
• Verification of minimum safety distances and guard positioning according to ISO 13855 and ISO 13857
• Environmental and robustness testing, including electromagnetic compatibility (EMC), temperature, vibration, and power supply interruption behavior
• Software and configuration review, including parameter verification, safety function configuration management, and change control

Testing Methods

Power and Force Limiting Validation

Force and pressure testing is performed using calibrated load cells, force gauges, pressure-indicating film, and pressure sensor matrices applied at realistic contact locations across the human body regions specified in ISO/TS 15066. Measurements distinguish between quasi-static contact, where a body part may be trapped or clamped, and transient contact, where the robot strikes the operator at speed. The measured peak forces and pressures are compared against the permissible biomechanical limits, and the maximum permissible robot speed at each contact location is derived from the test data. Where repeatable physical testing is impractical, validated simulation and analytical models may supplement measurements, provided the results are conservative.

Speed and Separation Monitoring Tests

Separation monitoring is evaluated by combining measured robot speed, measured stopping performance, and sensor detection response times into the protective separation distance calculation defined by ISO 13855 and ISO/TS 15066. Robot motion is tracked using laser trackers, motion capture systems, or internal safety encoders, while the detection performance of safety scanners or vision systems is verified at the boundaries of the protective field. Testing confirms that the robot decelerates or executes a protective stop before the operator can reach the moving robot.

Stop Performance and Safety Function Validation

Stopping time and stopping distance are measured with stop-time analyzers at worst-case conditions of payload, extension, speed, and tool orientation. Each safety function — protective stop, emergency stop, speed monitoring, space limiting, and monitored stop — is exercised through both functional testing and fault injection. Deliberate faults are introduced into sensors, logic, and output circuits to confirm that the safety control system behaves safely under single-fault conditions, as required for Category 3 architecture under ISO 13849-1.

Risk Assessment and Documentation Review

Engineers conduct a task-based risk assessment covering all identified hazards, estimate and evaluate risk for each hazard zone, and verify that the implemented protective measures match the risk reduction requirements. The assessment concludes with a validation report, including a traceability matrix linking each hazard to its protective measure and corresponding test evidence, which is essential for CE marking, third-party type examination, and regulatory acceptance.

Applications

HRC safety assessment is applied wherever collaborative robots share workspace with people, and demand is growing rapidly across labor-intensive and high-mix manufacturing environments:

• Automotive final assembly, door and dashboard installation, and bolt tightening alongside human operators
• Electronics and semiconductor manufacturing, including PCB handling, connector insertion, and precision assembly
• Medical device and pharmaceutical production, where cleanroom-compatible cobots assist with filling, inspection, and packaging
• Machine tending for injection molding machines, CNC centers, and press brakes
• Quality inspection and testing stations combining robot handling with human judgment
• Logistics and e-commerce fulfillment, including piece picking, palletizing, and co-packing
• Food and beverage handling with washdown-rated collaborative systems
• Laboratory automation and research institutions developing next-generation human-robot interaction
• Aerospace drilling, fastening, and composite layup assistance under strict process traceability

FAQ

What is the difference between a collaborative robot and a collaborative application?

A collaborative robot is a piece of hardware with inherent safety-oriented features such as force limiting and compliant joints. A collaborative application is a complete robot system deployed in a specific task and environment. ISO 10218-2 and ISO/TS 15066 require assessment at the application level: even a certified cobot can be unsafe if the end-effector is sharp, the workpiece is hot, or the process generates hazards that force limiting cannot address.

Which standards are mandatory for HRC safety assessment?

ISO 10218-1 and ISO 10218-2 form the primary basis, with the 2025 revisions incorporating the collaborative requirements of ISO/TS 15066. Risk assessment per ISO 12100 and functional safety verification per ISO 13849-1 or IEC 62061 are essential supporting elements. Regional regulations, such as the EU Machinery Regulation and OSHA requirements in the United States, determine how these standards apply for market access and workplace compliance.

How are force and pressure limits determined during testing?

ISO/TS 15066 Annex A defines maximum permissible force and pressure values for defined human body regions, based on biomechanical research on pain onset thresholds. Testing measures actual contact forces and pressures at each potential contact point and compares them against these limits, distinguishing quasi-static clamping contact from transient impact contact, which have different threshold values.

How long does a typical HRC safety assessment take?

For a single collaborative workcell, testing typically requires several days of on-site measurement plus one to three weeks for risk assessment, documentation review, and reporting. Complex multi-robot cells, mobile manipulators, or applications involving many contact scenarios may require several weeks or an extended program with staged validation during integration.

Is third-party certification required for collaborative robots?

Legal requirements vary by market. In the European Union, placing machinery on the market requires a conformity assessment and CE marking, and third-party type examination may be mandatory for certain machine categories. In many other jurisdictions, self-declaration supported by documented risk assessment and test evidence is acceptable, although third-party certification from an accredited body is widely used to reduce liability and support customer acceptance.

When should a collaborative application be reassessed?

Reassessment is required whenever the application changes, including a different end-effector, workpiece, speed, payload, layout, or operating procedure. Periodic revalidation, typically annually or according to the site safety management system, is also recommended because component wear, parameter drift, and software updates can degrade safety performance over time.