Power and Force Limiting (PFL) is a collaborative robot safety mode where the robot's inherent design or active control system restricts the kinetic energy and contact force of its movements to levels considered safe for incidental or expected contact with a human operator. This is a core requirement defined in the ISO/TS 15066 technical specification, which provides biomechanical limits for pain and injury thresholds. PFL enables true physical collaboration by allowing a human and robot to work in direct contact or close proximity without traditional safety fencing.
Glossary
Power and Force Limiting (PFL)

What is Power and Force Limiting (PFL)?
Power and Force Limiting (PFL) is a fundamental safety mode for collaborative robots.
Implementation relies on inherently safe design (e.g., lightweight structures, rounded edges, compliant actuators) and active control strategies like torque sensing at each joint. When a sensor detects an unexpected force or torque exceeding a predefined threshold—indicating contact—the control system triggers an immediate protective stop. This mode is distinct from Speed and Separation Monitoring (SSM), which aims to prevent contact altogether. PFL is essential for applications like hand-guiding, assembly assistance, and precision tasks requiring shared manipulation of a workpiece.
Key Characteristics of PFL Systems
Power and Force Limiting (PFL) is a foundational safety mode for collaborative robots, defined by ISO/TS 15066. Its characteristics ensure safe physical interaction by design.
Inherently Safe Design
PFL is achieved through inherently safe design principles, not just reactive software controls. This includes:
- Force-limited joints with back-drivable actuators and mechanical compliance.
- Rounded, padded surfaces and absence of pinch points to minimize injury risk during contact.
- Low effective mass and inertia, ensuring the robot cannot transfer hazardous energy upon impact. The design ensures safety is maintained even in the event of a control system failure.
Quantified Biomechanical Limits
PFL systems are governed by pain and injury thresholds defined in ISO/TS 15066. The standard provides maximum permissible values for:
- Transient contact: Short-duration, quasi-static contact (e.g., a bump). Limits are defined for different body regions (e.g., forehead, hand, shin).
- Quasi-static contact: Prolonged clamping or trapping. Force and pressure limits are significantly lower. These limits are based on biomechanical studies and are the absolute foundation for setting a robot's power and force parameters.
Dynamic Performance Monitoring
PFL is enforced in real-time by the robot's control system, which continuously monitors and limits:
- Joint torque and current: Directly correlates with applied force.
- Speed and momentum: Power is a function of force and velocity; limiting both is essential.
- Contact detection: Using built-in joint torque sensors or motor current measurements to detect unexpected external forces indicative of human contact. Upon detection, the robot must enter a protective stop or reduce its energy to safe levels.
Integration with Other Safety Modes
PFL is one of four collaborative operation safety modes per ISO 15066. It is often used in conjunction with:
- Hand Guiding: The robot is back-drivable under PFL, enabling safe physical guidance for programming.
- Speed and Separation Monitoring (SSM): PFL acts as a last line of defense if the protective separation distance is breached.
- Safety-Rated Monitored Stop: The robot stops before human entry; PFL enables safe resumption of work in close proximity. A true collaborative robot seamlessly transitions between these modes based on sensor input.
Validation and Risk Assessment
Deploying a robot in PFL mode requires a formal risk assessment and physical validation. This process involves:
- Identifying all potential contact scenarios (e.g., collision, clamping, guided movement).
- Measuring transmitted forces and pressures using a force-pressure measurement device as specified by the standard.
- Verifying that measured values are below the biomechanical limits for all tested points on the robot's surface and across its entire workspace. Compliance is not assumed; it must be empirically verified for the specific application.
Application-Specific Tuning
While based on fixed limits, PFL implementation is not one-size-fits-all. Key tuning parameters include:
- Payload and tooling: The end-effector mass and geometry drastically change collision dynamics. Limits must be validated with the full tool attached.
- Task and trajectory: High-speed motions or movements near obstacles require more conservative settings.
- Collaborative workspace layout: The proximity of humans dictates whether PFL is the primary or secondary safety layer. Proper tuning balances safety with operational efficiency.
PFL vs. Other Collaborative Safety Modes
This table compares the technical implementation, safety mechanisms, and operational characteristics of Power and Force Limiting (PFL) against the other primary collaborative safety modes defined by ISO/TS 15066.
| Feature / Metric | Power and Force Limiting (PFL) | Speed and Separation Monitoring (SSM) | Hand Guiding | Safety-Rated Monitored Stop |
|---|---|---|---|---|
Primary Safety Mechanism | Inherently limited joint torque and power | Maintains protective separation distance via sensors | Direct physical control by human operator | Robot stops motion before human enters workspace |
Physical Contact with Human | Allowed (incidental or intentional) | Not allowed; must stop before contact | Required for operation | Not allowed; robot must be stationary |
Required External Safeguarding | Minimal to none (inherent safety) | Perimeter laser scanners, safety mats, or vision systems | Safety-rated enabling device on the teach pendant | Traditional hard guarding (e.g., light curtains, fences) when not in stop mode |
Robot Motion During Collaboration | Continuous operation at reduced force | Continuous operation at speed controlled by separation | Motion only when directly guided by human | No motion; robot is in a safe stopped state |
Typical Force/Power Limit | < 150 N (industry standard threshold) | N/A (contact prevention) | Defined by human operator's applied force | N/A |
Maximum Allowable Speed | Speed limited to keep power/force below threshold | Speed dynamically reduced as separation distance decreases | Speed limited by control system, typically slow | N/A (stopped) |
ISO/TS 15066 Transient Contact Metrics | Defined for 29 body regions (e.g., forehead: 190 N max) | N/A | N/A | N/A |
Operator Skill Level Required | Low (intuitive co-existence) | Low (system manages separation) | Moderate (requires training for path teaching) | Low (simple start/stop interaction) |
Best Suited For | Assembly, machine tending, direct hand-over tasks | Material handling, packaging where human moves around cell | Path programming, complex trajectory teaching | Infrequent access for maintenance or part loading |
Common Applications of PFL
Power and Force Limiting (PFL) enables safe physical collaboration by restricting a robot's kinetic energy. Its applications span industries where human dexterity and machine strength must work in close proximity.
Assembly and Kitting
PFL cobots are integral to light assembly tasks where a human and robot work on the same component. The robot can handle repetitive, precise operations like screw driving, part insertion, or applying adhesive, while the human performs complex wiring or final inspection. The force limits ensure that incidental contact during hand-offs or close-quarters work does not cause injury.
- Example: A cobot presents a circuit board while a worker manually attaches connectors.
- Key Benefit: Combines human flexibility with robotic consistency without safety fencing.
Machine Tending and CNC
In machine tending, a PFL-equipped robot loads and unparts parts from CNC machines, lathes, or injection molding machines. The human operator can safely enter the shared workspace to perform quality checks, clear jams, or change tooling while the robot is operational. The robot's inherently limited joint torque prevents crushing injuries if a worker's hand is near the gripper or a part.
- Example: A cobot retrieves a machined component, allowing a technician to measure critical dimensions in-cycle.
- Key Benefit: Maximizes equipment uptime by eliminating full safety shutdowns for human intervention.
Packaging and Palletizing
PFL robots handle final packaging, case packing, and low-speed palletizing in environments where logistics are dynamic. They can collaborate with human packers who place irregular items or perform last-minute customizations. The force limits are crucial when handling potentially fragile items or when human hands are frequently in the robot's operational envelope to adjust packaging or apply labels.
- Example: A cobot places filled boxes onto a pallet while a worker inserts promotional leaflets.
- Key Benefit: Enables flexible, low-volume, high-mix production lines without extensive reconfiguration.
Quality Inspection and Testing
Cobots in PFL mode are used to present parts to human inspectors or to perform non-destructive testing requiring human guidance. The robot can maneuver a component under a camera or sensor, while the inspector can physically adjust the part's orientation for a better view. The compliant control allows the human to easily guide the robot's arm, a form of kinesthetic teaching, to define new inspection points.
- Example: A worker guides a cobot-mounted camera to inspect weld seams on a large, irregular assembly.
- Key Benefit: Enhances inspection thoroughness by combining robotic precision with human visual judgment.
Laboratory Automation and Biosciences
In research labs and pharmaceutical settings, PFL cobots automate delicate tasks like pipetting, plate handling, and instrument loading. Scientists can work adjacent to the robot, preparing samples or reagents, with the assurance that accidental contact will not break expensive glassware or cause harm. The smooth, force-limited motion is essential for handling biological samples without agitation.
- Example: A cobot transfers microplates between an incubator and a reader while a technician adds reagents.
- Key Benefit: Increases experimental throughput and reproducibility while maintaining a safe, interactive lab environment.
Finishing and Polishing
Applications like sanding, deburring, and polishing often require the tool (e.g., a sanding pad) to be in continuous contact with a workpiece with controlled force. A PFL cobot can perform the repetitive motion, while a human supervisor can frequently intervene to feel the surface quality, change abrasives, or hold the workpiece. The robot's impedance control (often underlying PFL) allows it to maintain a constant contact force safely.
- Example: A cobot sands a composite aerospace part; a worker periodically feels the surface and adjusts the robot's path.
- Key Benefit: Automates ergonomically taxing tasks while allowing for expert human oversight and adjustment.
Frequently Asked Questions
Power and Force Limiting (PFL) is a foundational safety standard for collaborative robots. These FAQs address its technical implementation, standards, and role in modern human-robot interaction.
Power and Force Limiting (PFL) is a collaborative robot safety mode where the robot's inherent mechanical design or real-time control software actively caps the kinetic energy and contact force of its movements to levels deemed safe for incidental human contact. It works through a combination of inherently safe design (e.g., lightweight structures, rounded edges, joint torque sensors) and active control. The control system continuously monitors joint torque and velocity. If a collision is detected—typically by a measured torque exceeding a predefined threshold—the robot executes a protective stop. This safety function is formally defined in the international standard ISO 10218-1 for industrial robots and detailed in the technical specification ISO/TS 15066 for collaborative operation.
Enabling Efficiency, Speed & Accuracy
Intelligent Analysis, Decision & Execution
We build AI systems for teams that need search across company data, workflow automation across tools, or AI features inside products and internal software.
Talk to Us
Search across company data
Give teams answers from docs, tickets, runbooks, and product data with sources and permissions.
Useful when people spend too long searching or get different answers from different systems.

Automate internal workflows
Use AI to route work, draft outputs, trigger actions, and keep approvals and logs in place.
Useful when repetitive work moves across multiple tools and teams.

Add AI to products and internal tools
Build assistants, guided actions, or decision support into the software your team or customers already use.
Useful when AI needs to be part of the product, not a separate tool.
Related Terms
Power and Force Limiting (PFL) is one of four defined collaborative operation modes. These related terms detail the other safety strategies, complementary technologies, and core standards that define safe human-robot collaboration.
Speed and Separation Monitoring (SSM)
A collaborative robot safety mode where the robot's speed is dynamically controlled to maintain a calculated protective separation distance from a human, ensuring the robot can stop before contact occurs. This relies on real-time perception systems like LiDAR or 3D cameras to track the human's position and velocity.
- Core Principle: Maintain a minimum safe distance that accounts for robot stopping time and human intrusion speed.
- Contrast with PFL: SSM is designed to prevent contact, whereas PFL is designed for safe incidental contact.
- Application: Ideal for tasks where the robot and human work in adjacent but non-overlapping workspaces.
Collaborative Robot (Cobot)
A robot designed from the ground up to operate safely alongside humans in a shared workspace. PFL is a core enabling technology for most modern cobots.
- Inherent Safety Features: Include force-limited joints, rounded edges, absence of pinch points, and smooth surfaces.
- Control Integration: PFL is typically implemented through series elastic actuators or torque sensors in the joints, coupled with software that limits motor current.
- Market Impact: Cobots have democratized automation by allowing flexible, close-proximity work in SMEs and complex assembly lines.
Physical Human-Robot Interaction (pHRI)
The subfield of HRI concerned with direct physical contact and force exchange between a human and a robot. PFL is a foundational safety requirement for enabling pHRI.
- Research Focus: Includes impedance control, admittance control, and physical collaboration where the human and robot jointly manipulate an object.
- Applications: Physical rehabilitation, cooperative assembly, and hand-over tasks.
- Safety-Critical: Demands rigorous validation against biomechanical standards like ISO/TS 15066 to prevent injury during intended or unintended contact.
Hand Guiding
A collaborative operation mode where a human operator physically grasps a robot's end-effector or dedicated handle and directly guides its motion. The robot's control system detects the applied force and moves compliantly.
- Activation: Typically requires a dedicated, safety-rated enabling device or button that must be continuously pressed.
- Programming Use: A primary method for kinesthetic teaching, allowing rapid task demonstration without code.
- Relationship to PFL: Often relies on the same underlying force/torque sensing and compliant control infrastructure as PFL to ensure safe, responsive guidance.
Safety-Rated Monitored Stop
The simplest collaborative operation mode. The robot comes to a complete, safety-monitored stop when a human enters the collaborative workspace, and only resumes when the human leaves.
- Function: The robot acts as a static obstacle until the space is clear. This mode does not involve concurrent motion.
- Implementation: Uses safety-rated sensors (light curtains, area scanners) to detect human presence and trigger a validated safe stop.
- Use Case: Suitable for tasks where the human only needs intermittent access to the robot's workspace, such as part loading/unloading.

About the author
Prasad Kumkar
CEO & MD, Inference Systems
Prasad Kumkar is the CEO & MD of Inference Systems and writes about AI systems architecture, LLM infrastructure, model serving, evaluation, and production deployment. Over 5+ years, he has worked across computer vision models, L5 autonomous vehicle systems, and LLM research, with a focus on taking complex AI ideas into real-world engineering systems.
His work and writing cover AI systems, large language models, AI agents, multimodal systems, autonomous systems, inference optimization, RAG, evaluation, and production AI engineering.
Partnered with leading AI, data, and software stack.
How We Work
Custom AI workflows for your Business
One-fit-all AI don't work for modern businesses. At Inferensys, we aim to understand your business & custom requirements; which we use to define most efficient agentic workflows, the data, and the tools for your business.
01
Review the use case
We understand the task, the users, and where AI can actually help.
Read more02
Pick the right approach
We define what needs search, automation, or product integration.
Read more03
Build the first useful version
We implement the part that proves the value first.
Read more04
Improve from there
We add the checks and visibility needed to keep it useful.
Read moreThe first call is a practical review of your use case and the right next step.
Talk to Us