Robot Standard

Type	Class
Resource	urn:ngm:class:robot-standard
Domain	robotics
Status	stable
Maturity	established
Quality	0.52
Is a	Safety and Standards Technical Standard Safety Regulation Normative Framework Robotics Industrial Safety Product Regulation
Has part	Conformity Assessment IEEE 1872 ISO 10218 ISO 13482 ISO 3691-4 ISO TS 15066 Safety Function Test Method Risk Assessment OPC UA Robotics Performance Criterion Normative Annex
Requires	CE Marking Conformity Assessment Body Declaration of Conformity Harmonised Standard Protective Measure Safety Function Technical Documentation Risk Assessment Protective Measure
Enables	CE Marking Cobot Deployment Digital Product Passport IIoT Integration Market Access Safe Human-Robot Collaboration Safe Human-Robot Collaboration Robot Interoperability Autonomous Mobile Robot Safety Robot Skill Transfer IIoT Integration
Generated	by: process:vault-migrate/1.0 · at: 2026-09-22T12:41:39.024503178Z
Contrasts with	De Facto Standard Open Source Middleware Self-Certification Domain-Specific Framework ROS Industrial Voluntary Guideline
Depends on	Machinery Directive Functional Safety IEC 61508 IEC 62061 ISO 13849 Product Liability EU AI Act Regulatory Instrument NIS2 Directive Machinery Regulation
Implemented in layer	SystemDesignLayer ManufacturingLayer IntegrationLayer ComplianceLayer OntologyLayer
Implements	ISO/TC 299 OWL Ontology Performance Level Method Safety Integrity Level Speed and Separation Monitoring Force Control EU Machinery Regulation EHSR Framework OWL Ontology OPC UA NodeSet Speed and Separation Monitoring Performance Level Method
Legacy term id	RB-9027
Links	AMRC ANSI Autonomous Mobile Robot Safety Autonomous Mobile Robots BARA BSI CE Marking Cobot Deployment Collaborative Robots ComplianceLayer Conformity Assessment Conformity Assessment Body De Facto Standard Declaration of Conformity Digital Product Passport DIN Domain-Specific Framework Driverless Industrial Trucks EHSR Framework EU Machinery Regulation European Commission Exoskeletons Force Torque Sensing Harmonised Standard IEC 61508 IEC 62061 IEC 62443 IEC 80601-2-78 IEEE 1872 IEEE Robotics and Automation Society IIoT Integration Industrial Robots Industrial Safety IntegrationLayer ISO 10218 ISO 12100 ISO 13482 ISO 13849 ISO 13855 ISO 22166 ISO 25132 ISO 3691-4 ISO 8373 ISO 9283 ISO SAE 21434 ISO TC 299 ISO TS 15066 Machinery Directive Machinery Regulation ManufacturingDomain ManufacturingLayer Market Access Modular Robots MTC NIOSH NIOSH Guidelines NIS2 Directive Normative Annex Normative Framework NPL Ontology Engineering OntologyLayer OPC Foundation OPC UA OPC UA NodeSet OPC UA Robotics Open Source Middleware OWL 2 DL OWL Ontology Performance Criterion Performance Level Performance Level Method Personal Care Robots Product Liability Product Regulation Protective Measure RegulatoryComplianceDomain RIA Risk Graph Method Robot Interoperability Robot Skill Transfer RoboticsGroundedDomain ROS Industrial Safe Human-Robot Collaboration SafetyEngineeringDomain Safety Function Safety Integrity Level Safety Regulation Self-Certification SLAM Navigation Speed and Separation Monitoring Surgical Robots SystemDesignLayer Technical Documentation Test Method Time-Sensitive Networking University of Manchester Voluntary Guideline EU AI Act Force Control Functional Safety Human-Robot Interaction Laser Scanner Machine Learning Risk Assessment Robotics Swarm Robotics Technical Standard
Page resource	urn:visionflow:page:d0c3da0c5e9a4ae59d135908dfb2f62fb2ac2da4497d22fb4682b392b1dcc867
Related to	IEC 61508 IEC 62443 ISO 12100 ISO 13849 ISO 8373 ISO 9283 ISO SAE 21434 ROS-Industrial EU AI Act Regulatory Instrument NIOSH Guidelines IEC 80601-2-78 ISO 13855 ISO 10218 ISO TS 15066 ISO 13482 ISO 3691-4 ISO 22166 ISO 25132 IEEE 1872 OPC UA Robotics EU Machinery Regulation IEC 62061 Robotics Collaborative Robots Autonomous Mobile Robots Industrial Safety Functional Safety Ontology Engineering Human Robot Interaction Machine Learning Discipline Risk Assessment Performance Level Safety Integrity Level BSI NPL AMRC University of Manchester MTC BARA ISO TC 299 IEEE Robotics and Automation Society RIA OPC Foundation NIOSH
Standardized by	ANSI BSI European Commission IEEE Robotics And Automation Society ISO/TC 299 OPC Foundation RIA BSI DIN
Supports	Autonomous Mobile Robots Collaborative Robot Exoskeleton Industrial Robot Surgical Robot Swarm Robotics Collaborative Robots Industrial Robots Personal Care Robots Exoskeletons Surgical Robots Driverless Industrial Trucks Modular Robots
Uses	Force Torque Sensing Ontology Engineering OPC UA OWL 2 DL Performance Level Safety Integrity Level Time-Sensitive Networking Laser Scanner Risk Graph Method Performance Level Force Torque Sensing SLAM Navigation OWL 2 DL

RobotStandard is a normative technical instrument — a specification, guideline, or regulation — issued by a recognised standards development organisation (SDO) or regulatory body to define mandatory or voluntary safety requirements, performance criteria, interoperability protocols, and terminolog…

Semantic Classification

Content

Compositional Relationships (Components)

SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:hasPart rb:SafetyRequirement))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:hasPart rb:RiskAssessmentProcedure))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:hasPart rb:ConformityAssessmentRoute))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:hasPart rb:PerformanceCriterion))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:hasPart rb:TerminologyClause))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:hasPart rb:TestMethod))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:hasPart rb:NormativeAnnex))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:hasPart rb:InformativeGuidance))

## Dependency Relationships
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:requires rb:RiskAssessment))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:requires rb:ConformityAssessmentBody))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:requires rb:HarmonisedStandard))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:requires rb:CEMarking))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:requires rb:TechnicalDocumentation))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:dependsOn rb:FunctionalSafety))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:dependsOn rb:IEC61508))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:dependsOn rb:ISO13849))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:dependsOn rb:MachineryDirective))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:dependsOn rb:ProductLiability))

## Capability Relationships
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:enables rb:SafeHumanRobotCollaboration))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:enables rb:RobotInteroperability))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:enables rb:CEMarking))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:enables rb:MarketAccess))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:enables rb:CobotDeployment))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:enables rb:RobotSkillTransfer))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:supports rb:CollaborativeRobots))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:supports rb:AutonomousMobileRobots))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:supports rb:PersonalCareRobots))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:supports rb:DriverlessIndustrialTrucks))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:supports rb:SurgicalRobots))

## Implementation Relationships
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:implements rb:ISOTechnicalCommittee299))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:implements rb:EHSRFramework))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:implements rb:OWLOntology))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:implements rb:OPCUANodeSet))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:implements rb:ForceTorqueLimit))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:implements rb:SpeedSeparationMonitoring))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:uses rb:PerformanceLevelMethod))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:uses rb:SafetyIntegrityLevel))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:uses rb:RiskGraphMethod))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:uses rb:FaultTreeAnalysis))

## Reduction Relationships
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:reduces rb:WorkplaceInjuryRisk))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:reduces rb:IntegrationComplexity))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:reduces rb:RegulatoryFragmentation))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:reduces rb:LiabilityExposure))
SubClassOf(rb:RobotStandard
  ObjectSomeValuesFrom(rb:reduces rb:TimeToMarket))

## Data Properties (Characteristics)
DataPropertyAssertion(rb:hasIdentifier rb:RobotStandard "RB-9027"^^xsd:string)
DataPropertyAssertion(rb:authorityScore rb:RobotStandard "0.87"^^xsd:decimal)
DataPropertyAssertion(rb:globalRobotStandardsBodyCount rb:RobotStandard "6"^^xsd:integer)
DataPropertyAssertion(rb:iso10218PublicationYear rb:RobotStandard "2011"^^xsd:integer)
DataPropertyAssertion(rb:iso15066PublicationYear rb:RobotStandard "2016"^^xsd:integer)
DataPropertyAssertion(rb:iso13482RevisionYear rb:RobotStandard "2024"^^xsd:integer)
DataPropertyAssertion(rb:machineryRegulationEffectiveYear rb:RobotStandard "2027"^^xsd:integer)
DataPropertyAssertion(rb:opcuaRoboticsMarketSharePct rb:RobotStandard "78"^^xsd:integer)
DataPropertyAssertion(rb:globalRobotShipmentsPerYear rb:RobotStandard "130000"^^xsd:integer)
DataPropertyAssertion(rb:biomechanicalBodyRegionPairs rb:RobotStandard "29"^^xsd:integer)
DataPropertyAssertion(rb:plfQuasistaticForceExampleN rb:RobotStandard "130"^^xsd:integer)
DataPropertyAssertion(rb:ssmHandApproachSpeedMmPerS rb:RobotStandard "2000"^^xsd:integer)
DataPropertyAssertion(rb:amrEmergencyStopResponseS rb:RobotStandard "0.5"^^xsd:decimal)

## Property Constraints
SubClassOf(rb:RobotStandard
  DataAllValuesFrom(rb:isNormative xsd:boolean))
SubClassOf(rb:RobotStandard
  DataSomeValuesFrom(rb:hasVersionNumber xsd:string))
SubClassOf(rb:RobotStandard
  DataMinCardinality(1 rb:hasScopeClause xsd:string))
SubClassOf(rb:RobotStandard
  DataMinCardinality(1 rb:hasConformityAssessmentRoute xsd:string))
SubClassOf(rb:RobotStandard
  DataMaxCardinality(1 rb:hasPublicationDate xsd:date))

## Annotations
AnnotationAssertion(rdfs:label rb:RobotStandard "Robot Standard"@en)
AnnotationAssertion(rdfs:comment rb:RobotStandard "Normative instrument defining safety, performance, interoperability, and terminology for robotic systems; structured around ISO TC 299, IEEE 1872 CORA ontology, OPC UA Robotics companion specification, RIA R15.06, and EU Machinery Regulation 2023/1230 effective 14 January 2027. Nine-tier standards hierarchy covering ISO 10218 industrial safety, ISO/TS 15066 collaborative robots, ISO 13482:2024 personal care robots, ISO 3691-4 Amd1:2024 driverless trucks, ISO 22166 modular robots, ISO 25132 skill ontology, IEEE 1872 CORA, OPC UA Robotics, and EU Machinery Regulation 2023/1230."@en)
AnnotationAssertion(dcterms:identifier rb:RobotStandard "RB-9027"^^xsd:string)
AnnotationAssertion(dcterms:subject rb:RobotStandard "Robotics, Safety Standards, Compliance, ISO TC 299, Collaborative Robots, Machinery Regulation, Ontology, OPC UA"@en)

)

Property Characteristics

AsymmetricObjectProperty(rb:requires) AsymmetricObjectProperty(rb:enables) AsymmetricObjectProperty(rb:implements) AsymmetricObjectProperty(rb:reduces) TransitiveObjectProperty(rb:dependsOn) FunctionalDataProperty(rb:hasVersionNumber) FunctionalDataProperty(rb:hasPublicationDate) FunctionalDataProperty(rb:authorityScore) FunctionalDataProperty(rb:machineryRegulationEffectiveYear)

Safety Function Performance Level Determination

The Performance Level (PL) determination methodology from ISO 13849-1 is the cornerstone of robot safety function specification. For each hazardous situation identified in the risk assessment, the safety function performance requirement is derived through the risk graph (ISO 13849-1 Annex A) considering severity of injury, frequency of exposure, and avoidance possibility.

PL Categories and Hardware Architectures:

Category 1: single-channel, well-tried components, no fault detection — suitable only for PLa and PLb (minor or reversible injuries, infrequent exposure). Example: single light curtain for low-speed machine entry detection.

Category 2: single-channel with test functionality (online diagnostics via test pulse) — achieves PLc or PLd depending on diagnostic coverage (DC). Example: OSSD-output safety laser scanner with self-test applied to robot perimeter guarding.

Category 3: dual-channel with common-cause failure exclusion (cross-monitoring, dissimilar technology where possible) — achieves PLd or PLe. This is the minimum architecture required for robot speed monitoring and joint torque limiting safety functions per ISO 10218-1 clause 5.4.

Category 4: dual-channel with high DC (≥99%) and mandatory cross-monitoring validated by on-demand testing — achieves PLe. Required for safety-critical functions where a single dangerous failure must not lead to loss of safety function.

Probability of Dangerous Failure per Hour (PFHd) targets:

PLa: PFHd ≥ 10⁻⁵ to < 10⁻⁴ (mean dangerous failure once per 1,000-10,000 hours) PLb: PFHd ≥ 3×10⁻⁶ to < 10⁻⁵ PLc: PFHd ≥ 10⁻⁶ to < 3×10⁻⁶ PLd: PFHd ≥ 10⁻⁷ to < 10⁻⁶ (mean dangerous failure once per 1M-10M hours) PLe: PFHd < 10⁻⁷

Safety Integrity Level (SIL) mapping: IEC 62061 and IEC 61508 SIL maps to ISO 13849-1 PL: SIL 1 ≈ PLc; SIL 2 ≈ PLd; SIL 3 ≈ PLe. Robot manufacturers may use either methodology; ISO 13849-1 is more prevalent in European cobot markets due to its mechanical-engineering-friendly graphical approach versus IEC 62061’s electrical safety engineering orientation.

ISO 10218 safety function PL requirements summary:

Safety-rated monitored stop (SRMS): PLd Category 3 minimum
Speed and separation monitoring (SSM): PLd Category 3, with protective device (safety laser scanner) certified to PLd
Power and force limiting (PFL): PLd Category 3, with joint torque sensing and force monitoring certified to PLd; validated against ISO/TS 15066 Annex A biomechanical limits
Hand guiding (HG): PLd Category 3 for the enabling device (three-position enable switch); force/torque interface to PLc minimum
Emergency stop function: PLd Category 3 per IEC 60204-1 stop category 0 or 1

ISO/TS 15066 SSM Protective Distance: Worked Calculation

The SSM protective distance formula — Ds = (Vh + Vr) × (Ts + Tc + Tr) + Zs + Zd + C — is demonstrated here for a representative cobot application.

Scenario: UR10e cobot operating in SSM mode alongside an electronics assembly operator. Safety laser scanner: SICK S300 Expert (response time Tc = 70 ms, detection capability 70 mm object at 4 m range, Zs = ±30 mm at 2 m, Zd = ±10 mm depth accuracy).

Parameter values:

Vh = 1,600 mm/s (conservative hand approach speed per ISO 13855, Annex C)
Vr = 250 mm/s (UR10e TCP speed in SSM zone, set in safety configuration)
Ts = 250 ms (UR10e Category 3 stop time from safety rated monitored stop function, measured per ISO 10218-2 clause 5.10.2)
Tc = 70 ms (SICK S300 Expert response time including OSSD switching)
Tr = 10 ms (PLC or safety controller response time to OSSD signal)
Zs = 30 mm (position measurement uncertainty of safety laser scanner at operating distance)
Zd = 0 mm (2D scanner, no depth axis; set to 0 or add separate Zd for 3D sensor)
C = 135 mm (intrusion depth per ISO 13855 Table 1, for 70 mm detection capability)

Calculation: Ds = (1,600 + 250) mm/s × (250 + 70 + 10)/1000 s + 30 + 0 + 135 mm Ds = 1,850 × 0.330 + 165 Ds = 610.5 + 165 Ds = 775.5 mm ≈ 780 mm minimum protective distance

This means the safety laser scanner protective field boundary must be set at least 780 mm from the robot’s nearest reachable position (considering maximum extension and any additional safety margins for workcell layout tolerances). Any operator presence detected within this zone triggers a Category 3 stop.

SSM zone design in practice: modern SSM implementations use dynamic zones that shrink as the robot slows (since Ts decreases at lower Vr), reducing unnecessary production stops whilst maintaining the same injury-prevention guarantee. Universal Robots SafeMove2 and KUKA SION implement dynamic SSM zone adaptation based on current TCP speed in real time.

About Robot Standards

Robot Standards form the normative backbone of safe, interoperable, and commercially viable robotic deployment. They function simultaneously as design specifications, risk-management frameworks, market-access instruments (providing harmonised routes to CE marking, UKCA marking, or equivalent conformity declarations), and semantic interoperability enablers defining ontologies and data exchange formats for IIoT architectures.
The prescriptive versus performance-based tension threads through every robot standard. Fixed industrial robot cells (KUKA KR QUANTEC body-in-white welding lines, FANUC R-2000 stamping presses) tolerate highly prescriptive safety-fence requirements because layouts are stable and hazard profiles are well-characterised.
By contrast, an AI-navigating nurse-assist cobot in an NHS ward, a rehabilitation exoskeleton learning a patient’s gait pattern, or a warehouse AMR fleet dynamically rerouting around pedestrians requires performance-based, context-adaptive safety assurance that rigid prescriptive clauses cannot anticipate.
The standards community has responded through a graduated move toward outcome-based requirements supported by formal methods, verified software, and run-time safety monitors — a shift that ISO/TS 15066’s biomechanical force limits represent imperfectly and that EU Machinery Regulation 2023/1230’s new AI-specific EHSRs attempt to formalise into law.
A secondary tension exists between Type C standards (product-specific, most prescriptive, directly usable for conformity assessment — ISO/TS 15066, ISO 13482, ISO 3691-4) and Type A/B standards (ISO 12100 risk assessment methodology; ISO 13849-1 control system safety; IEC 62061 functional safety) forming the normative underpinning.
Compliance with a Type C standard, when it exists, provides strong presumption of conformity with EHSRs; where no Type C standard exists (as for many AI-adaptive robot behaviours), the integrator must construct a safety case from Type A/B methods, requiring significantly greater engineering judgement and notified body engagement.

Standards Type Hierarchy

Type A — Generic Safety Principles: ISO 12100 (Risk assessment and risk reduction) is the foundational Type A standard underpinning all machinery safety including robotics. Its risk assessment methodology — hazard identification, risk estimation (severity × probability × exposure × avoidability), and risk reduction hierarchy (inherently safe design → safeguarding → information for use) — is referenced normatively by ISO 10218-1, ISO 10218-2, ISO 13482, and ISO 3691-4.
Type B1 — Safety Aspects: ISO 13849-1 (Safety-related parts of control systems) provides the Performance Level (PL) methodology for quantifying the risk reduction capability of safety functions. Five PLs (a through e) correlate with required probability of dangerous failure per hour: PLa: ≥10⁻⁵ to <10⁻⁴; PLd: ≥10⁻⁷ to <10⁻⁶; PLe: <10⁻⁷.
For robotics: safety-rated monitored stop, speed monitoring, and force limiting functions are typically required at PLd Category 3 (single-channel with monitoring, diagnostic coverage DC ≥ 90%). ISO 13849-1 is the dominant safety function specification methodology for robot systems in Europe and adopted by RIA R15.06 in North America.
Type B2 — Safeguards: ISO 13855 (Positioning of safeguards with respect to approach speeds of parts of the human body) provides the calculation methodology for minimum separation distances between hazardous robot motion and presence-sensing safeguards (light curtains, safety laser scanners, area scanners).
The formula Ds = K × (Ts + Tc + Tr) + C — where K is hand approach speed (2,000 mm/s for vertical/horizontal light curtains), Ts is robot stopping time, Tc is safeguard response time, Tr is any additional response time, and C is intrusion depth based on resolution — is the primary engineering input to robot workcell layout design and directly cited in ISO 10218-2 clause 5.4.
Type C — Specific Machine Standards: ISO 10218-1, ISO 10218-2, ISO/TS 15066, ISO 13482, and ISO 3691-4 are all Type C standards. They incorporate Type A/B methodology by reference and add robot-type-specific requirements: kinematic configurations to assess, specific safety functions required, test procedures for validation, and technical file documentation requirements.

Risk Assessment Framework Applied to Robotics

ISO 12100 risk assessment for a robot system proceeds through a structured sequence producing documentation forming part of the technical file required for EU Machinery Directive/Regulation and UK Supply of Machinery Regulations compliance.
Step 1 — Machine Limits Determination: define the robot system’s use limits (intended use, foreseeable misuse), space limits (safeguarded space, restricted space, operating space per ISO 10218-2 definitions), time limits (machine lifetime, service intervals, duty cycles), and other limits (maximum payload, maximum speed, environmental conditions including EMC environment, temperature range, contamination class).
Step 2 — Hazard Identification: systematic enumeration of hazards associated with the robot system. Mechanical hazards: crushing from robot motion (typically the primary hazard class); impact from high-velocity end-of-arm tooling; shearing from pinch points between robot arm segments and fixed structures. Electrical hazards: drive cabinet live parts access, EMI from robot drives affecting safety-rated sensors. Thermal hazards: weld spatter, hot workpiece handling, continuous-duty motor heating. Noise/vibration hazards: pneumatic tool impulse noise >85 dB(A), repetitive shock vibration. Ergonomic hazards: awkward operator posture during programming, cognitive load from human-robot task sharing. ISO 10218-2 Annex A provides a non-exhaustive significant hazards list tailored to robot systems.
Step 3 — Risk Estimation: for each hazard, estimate risk as a function of severity of harm (S1: reversible injury; S2: irreversible injury or death), frequency of exposure (F1: seldom; F2: frequent to continuous), probability of occurrence (P1: low; P2: high), and possibility of avoidance (A1: possible under specific conditions; A2: barely possible). The risk graph method (ISO 13849-1 Annex A) maps these parameters to required PL for each safeguarding function.
Step 4 — Risk Reduction: design solutions working down the hierarchy: (1) inherently safe design (software-imposed TCP speed limits, payload reduction to minimise kinetic energy, pinch point elimination by geometry); (2) safeguards (perimeter guarding per ISO 14120, ESPE light curtains, safety-rated speed monitoring per ISO 10218-2, collaborative operation modes per ISO/TS 15066); (3) information for use (residual risk warnings, operator training requirements, lockout/tagout procedures for maintenance access).
Step 5 — Documentation: technical file recording risk assessment results, safety function specifications (PL/SIL requirements and architectural solutions), validation test results, and residual risk information supporting the Declaration of Conformity and CE/UKCA marking.

IEEE 1872 CORA Formal Ontology Structure

The CORA (Core Ontology for Robotics and Automation) OWL 2 DL module defines the following principal class hierarchy and object property network, directly informing how ISO 25132 skill ontology constructs and OPC UA Robotics NodeClasses map to foundational knowledge representation concepts.

Top-Level CORA Classes (OWL declarations):

Robot rdfs:subClassOf PhysicalAgent
PhysicalAgent rdfs:subClassOf Agent
Agent rdfs:subClassOf PhysicalObject (for physical agents); or InformationalEntity (for software agents)
Actuator rdfs:subClassOf Component
Sensor rdfs:subClassOf Component
Component rdfs:subClassOf PhysicalObject
Platform rdfs:subClassOf PhysicalObject
Task rdfs:subClassOf Process
Action rdfs:subClassOf Process
Motion rdfs:subClassOf Action

CORA Object Properties (selected):
hasActuator: Robot → Actuator (ObjectProperty, inverse functionalProperty for arm endpoint)
hasSensor: Robot → Sensor
hasPlatform: Robot → Platform (functional)
performsTask: Agent → Task
executesAction: Robot → Action
hasPreCondition: Task → Condition
hasEffect: Task → Condition
locatedAt: PhysicalObject → Location (connects to DOLCE SpaceRegion)

CORA-to-ISO 25132 alignment (via CORA-Skills module, IEEE 1872-2025):
cora:Action aligns with iso25132:SkillClass (EquivalentClass under assumption of executable robot action)
cora:hasPreCondition aligns with iso25132:SkillPreCondition (SubPropertyOf)
cora:hasEffect aligns with iso25132:SkillPostCondition (SubPropertyOf)
cora:performsTask aligns with iso25132:hasSkillExecution for assembled skill composition plans

CORA-to-OPC UA Robotics alignment (via OPC UA Part 17 NodeSet mapping):
cora:Robot maps to opcua-robotics:MotionDevice (NodeClass)
cora:Actuator maps to opcua-robotics:Axes (individual joint axis NodeClass)
cora:Task maps to opcua-robotics:TaskControls (program execution state machine)
cora:Platform maps to opcua-robotics:MotionDeviceSystem (complete robot topology)

Components/Architecture of the Robot Standards Ecosystem

ISO 10218: Industrial Robot Safety (Tier 1)

ISO 10218-1:2011 (Safety requirements for industrial robots — Part 1: Robots) specifies requirements for the design and construction of industrial robots — the manipulator arm plus its dedicated control system.
Critical clauses include: control functions with safety characteristics (clause 5.4, mandating PLd Category 3 or SIL 2 minimum for any safety-rated monitoring function); speed and force limiting for single-axis and multi-axis slowing; end-of-arm tooling mechanical interface safety (maximum allowable coupling force, anti-drop provisions); control pendant safety (enabling device, three-position enable, dead-man release); and safety-rated monitored stop (SRMS) performance requirements.
The standard is undergoing revision as ISO 10218-1:2025 (FDIS stage early 2025), introducing: explicit safety-rated soft axes and Cartesian spaces as a functional safety layer enabling spatial restriction without physical guarding; updated alignment with ISO 13849-1:2023 and IEC 62061:2021; and provisions for collaborative operating modes moved from informative annexes into normative clauses.
ISO 10218-2:2011 (Part 2: Robot systems and integration) addresses the complete workcell — robot plus tooling, fixtures, guarding, control interfaces, and human operational tasks.
Critical additions include: safeguarded space design (clause 5.4 — presence-sensing device selection from ISO 13855, safety-rated boundary specification with minimum separation distance calculations based on hand speed K = 2,000 mm/s, overall system stopping performance Ts + Tc); all four collaborative operating modes (safety-rated monitored stop, hand-guiding, speed and separation monitoring, power and force limiting); commissioning and validation testing with documented pass/fail criteria; and the normative reference to ISO 12100 risk assessment methodology.
The upcoming ISO 10218-2:2025 substantially restructures collaborative operation annexes to align with ISO/TS 15066:2016 content, introduces explicit provisions for mobile manipulators (AMR base + arm combinations), and adds a new clause on safe integration of machine vision systems as PLd presence-sensing devices.
RIA R15.06 is the ANSI-adopted US equivalent under the RIA/ANSI harmonisation agreement. ANSI/RIA R15.06-2012 aligns with ISO 10218:2011 and remains the primary US compliance reference. The in-development R15.06-2024 update tracks ISO 10218:2025. The RIA additionally publishes ANSI/RIA R15.08 (industrial mobile robots), R15.02 (robot modularity), and R15.04 (workcell safety assessment methods).

ISO/TS 15066: Collaborative Robots (Tier 2)

ISO/TS 15066:2016 (Collaborative robots) is a Type-C technical specification supplementing ISO 10218 specifically for the four collaborative operating modes: safety-rated monitored stop (SRMS), hand guiding (HG), speed and separation monitoring (SSM), and power and force limiting (PFL).
The specification’s most influential contribution is Annex A — Biomechanical Limits: a normative table of 29 body-region pairs specifying acceptable quasi-static contact force (N), dynamic (transient) contact force (N), and contact pressure (N/cm²) derived from volunteer pain-threshold studies at Fraunhofer IFF Magdeburg (Behrens & Elkmann, 2012-2014) on adult male subjects aged 20-40.
Example Annex A entries: skull/forehead (130 N quasi-static, 130 N transient, 110 N/cm²); sternum (140 N / 210 N / 210 N/cm²); shoulder/deltoid (160 N / 210 N / 160 N/cm²); forearm/radius (160 N / 240 N / 180 N/cm²); ankle/foot (200 N / 250 N / 220 N/cm²).
PFL cobots using Annex A limits as their primary safety case for guarding removal include Universal Robots UR series (UR3e through UR30), FANUC CRX series, KUKA LBR iisy, ABB YuMi, and Doosan Robotics — together representing approximately 65% of collaborative robot unit shipments globally (IFR 2024).
SSM mode dynamic protective distance formula: Ds = (Vh + Vr) × (Ts + Tc + Tr) + Zs + Zd + C, where Vh is human body velocity component (measured or conservatively 1,600 mm/s for hand), Vr is maximum robot velocity at detected separation, Ts is robot stopping time, Tc is detection cycle time, Tr is response time, Zs is measurement uncertainty of sensing device, Zd is depth accuracy, and C is the intrusion depth constant from ISO 13855.
Safety laser scanners implementing SSM include: SICK S300 Expert, SICK S3000 (PLd-rated), Pilz PSENscan 3D, and Keyence SZ-V. Depth cameras (Intel RealSense D series, IFM O3D303 ToF cameras certified to PLd) are increasingly used for SSM enabling 3D operator body tracking.
Documented limitations: the Fraunhofer biomechanical data does not address paediatric, elderly, frail, or disabled populations (directly relevant to ISO 13482 personal care contexts); does not model cumulative low-force ergonomic exposure (NIOSH 2024-102 concern); and predates modern force/torque sensor technology at sub-Newton resolution that could enable finer safety strategies.
A revision to full IS (ISO 15066) was initiated in 2023, targeting updated biomechanical annexes with DLR/NIST/RIKEN demographic-expanded data, expanded SSM guidance for 3D LiDAR and depth cameras, explicit coverage of mobile cobot platforms, and contact modelling for simultaneous translation and rotation scenarios.
NIOSH Publication 2024-102 (provisional) supplements ISO/TS 15066 with occupational health guidance covering: ergonomic risk factors in cobot work (awkward posture, pace-forced collaborative assembly, psychosocial factors); noise and vibration from pneumatic end-effectors; chemical exposure from robotic welding fume in open-guarding configurations. Advisory rather than mandatory but increasingly cited in OSHA enforcement.

ISO 13482: Personal Care Robots (Tier 3)

ISO 13482:2014 covers three robot types: mobile servant robots (autonomous floor cleaning, delivery), physical assistant robots (lower-limb exoskeletons, transfer-assist), and person-carrier robots (autonomous wheelchairs).
The standard defines: risk assessment methodology for human-robot proximity in uncontrolled domestic and healthcare environments; contact-force limits for skin and joints consistent with ISO/TS 15066 Annex A; fall prevention requirements (mobile servant robots on inclines ≤6°, step transitions ≤15 mm); stability criteria under dynamic loading; and emergency stop performance (≤0.5 s from activation to complete stop).
ISO 13482:2024 (published December 2024) introduces major additions: explicit coverage of AI-driven behaviour uncertainty, requiring AI planning modules to be bounded by validated safe-operating envelopes verified through formal methods or systematic testing; updated force limits aligned with revised ISO/TS 15066 biomechanical data; a new data privacy clause referencing GDPR Article 9 (health data) for robots with cameras/microphones/biomedical sensors in domestic environments.
New in the 2024 revision: safety requirements for robots under remote oversight (teleoperation with latency compensation, defined minimum data rates for safe teleoperation at ≤200 ms round-trip latency); and alignment with IEC 80601-2-78 (medical beds) for transfer-assist robots in clinical settings, resolving a previously ambiguous interface between ISO 13482 and EU Medical Device Regulation.
Physical assistant robots/exoskeletons additionally fall under the emerging ISO 13482-3 (lower-limb exoskeletons for rehabilitation, additional neurological patient population provisions including fall-detection safety reflexes and muscle fatigue monitoring interfaces) and IEC 80601-2-78 for electro-medical equipment classification requiring IEC 60601-1 dielectric strength, leakage current, and EMC (IEC 60601-1-2) compliance alongside kinematic safety.

ISO 3691-4: Driverless Industrial Trucks (Tier 4)

ISO 3691-4:2020 (Driverless industrial trucks and their systems) covers automated guided vehicles (AGVs) using fixed infrastructure guidance and autonomous mobile robots (AMRs) using map-based SLAM navigation, deployed in warehouse, manufacturing, and logistics facilities.
The standard specifies: vehicle detection zones (warning zone, protective zone) and their relationship to speed and stopping performance; separation distance calculations for mixed-traffic zones; pedestrian crossing protection; load-handling interface safety; communication failure behaviour (fail-safe stop); and facility requirements for AMR operating areas.
Amendment 1 (ISO 3691-4:2020/Amd 1:2024) represents a major extension: enhanced requirements for AI-based obstacle detection systems, requiring learning-based perception to demonstrate performance equivalent to deterministic sensor systems; explicit performance requirements for SLAM-navigated AMRs requiring position accuracy certified to PLc (Category 2 hardware fault tolerance) or higher using certified safety laser scanners (SICK NAV3xx series, Pilz PITreader for zone management).
New in Amd 1:2024: provisions for multi-robot fleet coordination and traffic management arbitration (fleet manager fault behaviour must allow individual vehicle safe-stop without coordinated fleet halt); updated emergency stop response times (≤0.5 s from obstacle detection to complete stop, vehicles ≤1,500 kg at ≤2.5 m/s); V2X communication guidance for AMRs in mixed-mode facilities alongside human-driven forklifts using RFID/UWB transponders.
Amendment 1 is being implemented by Amazon Robotics (Proteus, Hercules, Sequoia systems), DHL Supply Chain, and Ocado Technology in their warehouse AMR fleet certification programmes.

ISO 22166: Modular Robots (Tier 5)

ISO 22166-1:2021 (Modular robots — Part 1: Vocabulary) establishes the definitional framework for modular reconfigurable robots: systems where joint modules, link modules, end-effector interface modules, and computing modules can be assembled into different kinematic configurations without factory-level recertification.
Key vocabulary: ModularUnit (atomic configurable element), KinematicInterface (standardised mechanical, electrical, and communication connection between modules), ConfigurationSpace (set of topological configurations achievable from a given module set), PerformanceClass (load capacity, reach, and speed category applicable to a configuration).
The companion ISO 22166-2 (safety and performance requirements, under development) addresses the primary commercialisation barrier: safety certification validity across configurations. Each new kinematic topology may require re-validation of dynamic models (mass matrix, gravity vector, Coriolis/centrifugal terms), safety-rated joint speed limits (dependent on effective payload lever arm, which changes with configuration), and collision geometry.
The emerging approach uses configuration-invariant safety properties — properties of individual modules (maximum joint torque, module structural integrity) that can be compositionally combined to bound the safety properties of any assembled configuration, avoiding the combinatorial explosion of per-configuration certification.
Modular robot manufacturers Hebi Robotics (X-series actuators), Robotis Dynamixel (servo modules), and Tomotor (industrial joint modules) are active contributors to ISO TC 299 WG7 developing this compositional certification approach.

ISO 25132: Robot Skill Ontology (Tier 6)

ISO 25132:2023 (Ontology for robot skill-based programming) defines a formal OWL 2-based ontology for describing robot skills — discrete, reusable, parameterisable capabilities composable into task plans and transferable between heterogeneous robot platforms.
Key ontology constructs: SkillClass (atomic executable capability, e.g. PickObject, PlaceObject, Weld, ScrewFasten, Inspect, Move); SkillParameter (contextual configuration values at instantiation, e.g. graspForce_N, targetPose_SE3, weldSpeed_mm_s); SkillPreCondition (logical assertions that must hold before execution, expressed as OWL class expressions or SPARQL ASK queries); SkillPostCondition (world state assertions guaranteed on successful completion, enabling STRIPS-style compositional task planning).
Additional constructs: SkillConstraint (parametric bounds enforced throughout execution by a runtime skill monitor); SkillComposition (sequential, parallel, and conditional composition operators for building task-level programs from skill primitives).
The skill ontology connects upward to IEEE 1872 CORA foundational concepts (Robot, Actuator, Task, Action, PhysicalObject) and downward to OPC UA Robotics for fieldbus-level execution interface binding, providing the semantic layer for a complete stack from abstract task specification to joint-space execution.
Programming by demonstration (PbD) workflows: an operator kinesthetically demonstrates a manipulation task; force/torque data and joint trajectories are recorded; a parameterised SkillClass description is extracted (trajectory segmentation, Gaussian Mixture Models); the skill is instantiated on a physically different robot using an ontology-based task planner generating robot-specific joint trajectories satisfying the abstract skill semantics.
Industrial applications include BMW Group Leipzig toolpath programming for adaptive deburring (2024), collaborative assembly skill transfer between KUKA LBR iiwa and UR10e for connector insertion, and cross-robot weld schedule transfer in tier-1 automotive supplier contexts.

IEEE 1872: CORA Ontology (Tier 7)

IEEE 1872-2015 (IEEE Standard Ontologies for Robotics and Automation) defines the Core Ontology for Robotics and Automation (CORA) using OWL 2 DL, providing foundational concepts for knowledge representation across heterogeneous robot systems.
CORA top-level taxonomy: Robot (artificial physical agent that can sense, reason, and act); Actuator (converts energy to physical action); Sensor (converts physical stimuli to information); Platform (physical base on which robot subsystems are assembled); PhysicalLocomotionCapability (intrinsic movement capabilities — translational, rotational, flying, swimming); Task (goal-oriented activity with preconditions, parameters, and intended effects); Action (atomic executable behaviour); Agent (entity capable of autonomous purposive action); PhysicalObject (any object in the robot’s environment that can be perceived, manipulated, or occupied).
CORA imports SUMO (Suggested Upper Merged Ontology, Niles & Pease 2001) and DOLCE (Descriptive Ontology for Linguistic and Cognitive Engineering, Gangemi et al. 2002) at the foundational upper level, providing formal grounding for spatio-temporal reasoning, event ontology, and role-based participation assertions.
A revision IEEE 1872-2025 completed first ballot round late 2024, extending CORA with four specialisation modules: CORA-AV (autonomous vehicles, V2X, traffic rule compliance); CORA-Swarm (swarm robotics, emergent collective behaviour, stigmergic coordination); CORA-AI (AI reasoning under uncertainty, sensor fusion, learned behaviour provenance aligned with W3C PROV-O, Bayesian belief revision); CORA-Skills (aligning CORA Task/Action taxonomy with ISO 25132 SkillClass hierarchy through formal axiom bridges).
CORA-AV, CORA-Swarm, and CORA-AI also add alignment with W3C Web of Things (WoT) Thing Descriptions for robot capability advertisement in IoT environments.

OPC UA Robotics Companion Specification (Tier 8)

OPC UA Part 17: Robotics (OPC 10000-17, Release 1.1.0, 2023) defines a standardised OPC UA information model and namespace for robot controllers, enabling MES, SCADA, and cloud analytics to discover, monitor, and control robots through a uniform semantic interface regardless of manufacturer.
NodeClasses defined: MotionDeviceSystem (complete robot mechanical and control topology); MotionDevice (individual manipulator, axis count, kinematic type, payload specification); Axes (per-axis joint type, position/velocity/effort limits, current actual values, homing state); TaskControls (program execution state machine — idle, running, suspended, error, emergency-stopped — mapped to ISO 10218-2 safety states); SafetyStates (ISO 10218-2 safety-rated operating modes mapped as OPC UA HasComponent references); LoadData (dynamic payload mass, CoM offset, inertia tensor for load monitoring).
Real-time data subscription through OPC UA PubSub (UADP encoding over UDP multicast) enables integration with TSN (Time-Sensitive Networking, IEEE 802.1Qbv) backbones for deterministic robot data streaming at ≤1 ms cycle time, enabling closed-loop machine vision and force control applications over standard Ethernet infrastructure.
Implementation status (2025): KUKA iiQKA OS 2.0+, Universal Robots PolyScope X, ABB OmniCore C30/C90, Yaskawa YRC1000 with OPC UA option, FANUC R-30iB Plus, Stäubli CS9, and Kawasaki E series — together covering approximately 78% of global industrial robot shipments (IFR 2024).
The ISO TC 299 / OPC Foundation Joint Working Group (established 2023) published JWG-01 (April 2025), mapping ISO 10218-2 safety states to OPC UA SafetyConditionClass instances for machine-readable safety state reporting in IIoT/Industry 5.0 architectures.

EU Machinery Regulation 2023/1230 (Tier 9)

Regulation (EU) 2023/1230 (the Machinery Regulation, published in the Official Journal of the EU, L 165, 29 June 2023) replaces Machinery Directive 2006/42/EC with direct regulatory effect from 14 January 2027. No transposition is required; the Regulation applies automatically in all EU member states. A 3.5-year preparation window (mid-2023 to January 2027) allows manufacturers to comply with either instrument during the transition.
Annex III EHSR 1.1.9 (Self-evolving machinery): the most significant innovation for robotics. Requires that any AI-induced behaviour change remain within pre-validated safe operating envelopes; manufacturers must implement online safety monitoring capable of detecting safety-envelope violations in real time; training data provenance, model validation, and post-deployment monitoring must form part of the technical file; and machinery must operate safely with degraded AI performance. Directly impacts AI-navigating cobots, AMRs with learned obstacle avoidance, and adaptive process robots.
Annex III EHSR 1.5.1 (Cybersecurity): first instance of cybersecurity as a mandatory EHSR in the EU machinery regime. Machines connected to external networks must implement security-by-design and document residual cybersecurity risks in the technical file. Connects to NIS2 Directive obligations and aligns with IEC 62443 series.
Annex III EHSR 2.1 (Autonomous machines): strengthened requirements for collision avoidance system performance, operator override authority (human must always be able to assert overriding authority over autonomous motion), and emergency stop accessible to persons in the vicinity (not just at fixed pendant locations).
Digital instructions: the Regulation mandates provision of operating instructions in digital format from 2027, enabling machine-readable instructions via OPC UA documentation node references and IIoT service delivery platforms.
The Commission’s AI-in-Machinery Expert Group (established 2024) published a preliminary guidance note in November 2024 mapping ISO/TS 15066 force limits as acceptable harmonised evidence for EHSR 2.1 compliance for PFL cobots, and ISO 3691-4 Amd1 as primary evidence for AMR EHSR 2.1 compliance. The delegated act on self-evolving machinery (EHSR 1.1.9) is expected as a Commission Delegated Regulation in Q2-Q3 2026.
UK UKCA divergence: post-Brexit UK machinery is governed by Supply of Machinery (Safety) Regulations 2008. The UK government intends to update these through statutory instrument, but has not confirmed a firm date matching EU 2027 nor whether the UK will adopt AI-specific EHSR provisions verbatim. UK manufacturers targeting both markets face potential double-conformity-assessment burden unless Mutual Recognition Agreement terms cover the updated machinery regimes.

EU Machinery Regulation 2023/1230: EHSR Technical Implementation Details

The self-evolving machinery EHSR (Annex III 1.1.9) introduces requirements that lack direct precedent in the machinery safety tradition and for which no harmonised standard yet exists as of May 2026. The following design approaches are being pursued by robot manufacturers to establish EHSR 1.1.9 compliance evidence:

Safe-Operating Envelope (SOE) Definition and Validation: The manufacturer must define a SOE as a formal set of constraints on robot states {q, dq, ddq, F, P} — joint positions, velocities, accelerations, contact forces, and Cartesian pose — that represent all physically reachable states under which the robot cannot injure a person even if the AI subsystem fails. The SOE must be validated through worst-case trajectory analysis, typically using reachability analysis tools (SpaceEx, CORA reachability toolbox, DynIBEX) proving that all reachable states under any AI-generated motion plan remain within the SOE.

Online Safety Monitor Architecture: The safety monitor runs in parallel with the AI motion planner on a safety-rated PLC (PLd Category 3) and continuously verifies: TCP velocity ≤ safe velocity limit for current separation distance; contact force at any joint ≤ ISO/TS 15066 Annex A limit for detected contact region; joint positions within safety-rated limits; and AI planning latency ≤ defined deadline (late outputs treated as planner failure, triggering safe stop).

AI Adaptation Bounds: Learning-enabled components must be constrained to adapt only parameters that cannot affect safety-relevant behaviour (e.g. approach speed beyond the safety-rated zone cannot be learned; only position within a pre-validated path can be refined). Adaptation history is logged to the technical file with timestamp, trigger condition, and parameter delta for post-incident analysis.

Cybersecurity Interface (EHSR 1.5.1): Robot controller firmware update mechanisms must implement: cryptographic signature verification of firmware packages (ECDSA-384 minimum); secure boot chain from hardware root of trust; network segmentation of robot control network from enterprise IT (OPC UA firewall or dedicated VLAN with IEC 62443 zone boundary enforcement); logging of all external command sessions with integrity-protected audit trail; and documented vulnerability disclosure and patching policy meeting the requirements of ISO 30162-1 (when published).

Conformity Assessment Routes in Practice

The route to CE marking under EU Machinery Directive 2006/42/EC (and from 2027, EU Machinery Regulation 2023/1230) depends on whether the robot system is listed in Annex IV (most industrial robots in cells with operator access — yes; stand-alone cobots in PFL mode — depends on force level and application) and whether harmonised standards covering all EHSRs are available and used.

Route 1 — Type Examination (Annex IX Directive / Annex VIII Regulation) by Notified Body: Required for Annex IV/Annex I high-risk machinery. The notified body (e.g. TÜV SÜD, TÜV Rheinland, SGS, Bureau Veritas, Intertek) examines the technical file, conducts representative type testing, and issues an EC/EU Type Examination Certificate. The manufacturer then enters production quality control (Annex X Directive) or full quality assurance (Annex XI). For AI-adaptive robots under the 2027 Regulation, notified bodies are expected to require both type testing of the base system and review of the adaptive safety monitoring architecture.

Route 2 — Full Quality Assurance (Annex XI Directive / Annex X Regulation): Manufacturer implements a quality management system certified by a notified body covering design, production, and testing. Less common for individual robot integrators due to QMS certification overhead; primarily used by major robot OEMs (KUKA, ABB, FANUC) for their standard robot product families.

Route 3 — Self-Certification (Annex VIII Directive / Annex VII Regulation): Available for non-Annex IV machinery where harmonised standards covering all EHSRs are used. The manufacturer prepares the technical file, applies harmonised standards (EN ISO 10218-2 + EN ISO 13849-1 + EN ISO 12100 + EN ISO 13855 + relevant Type C standards), issues the Declaration of Conformity, and affixes CE marking without notified body involvement. This route is used by most robot integrators for standard industrial workcells that do not include Annex IV machine functions.

UKCA marking (post-Brexit UK): equivalent to CE marking route under UK Supply of Machinery (Safety) Regulations 2008. UKCA Type Examination certificates must be issued by UK Approved Bodies (not EU notified bodies) for Annex IV equivalent machinery. Mutual recognition of CE and UKCA markings applies only where the UK government has specifically designated mutual recognition of particular conformity assessment procedures — a policy still under negotiation as of 2026.

Use Cases / Major Families of Robot Standards Deployment

Automotive body-in-white (BIW) welding cells: ISO 10218-2 is the primary standard for fixed industrial robot cells. A typical BIW line uses 200-400 articulated robots behind perimeter guarding per ISO 14120 with ESPE light curtains at operator access points meeting ISO 13855 separation distance requirements.
Risk assessment under ISO 10218-2 clause 5.4 drives guard height, access door interlocking (ISO 14119), and pneumatic energy isolation. CE marking follows the conformity assessment route through EN ISO 10218-2 + EN ISO 13849-1. The automotive sector drives approximately 130,000 welding robot installations per year globally (IFR 2024), approximately 60% in BIW applications.
Electronics assembly and semiconductor handling: cobots operating in ISO/TS 15066 PFL mode are widely deployed in electronics SMT inspection, cable harness routing, and small component assembly (smartphone camera module alignment requiring ±0.02 mm placement accuracy, USB-C connector insertion).
Universal Robots UR10e and UR20 in PFL mode achieve PLd (ISO 13849-1 Category 3) or SIL 2 (IEC 62061) for safety-rated force/torque monitoring via internal certified safety electronics (IFA certification for UR e-Series safety functions). Ergonomic redesign per NIOSH 2024-102 is applied when cobots replace high-repetition manual sub-assembly tasks.
Hospital logistics and pharmacy automation AMRs: autonomous mobile robots operating in clinical corridors (Aethon TUG for linen/medication transport, Swisslog TransCar for sterile goods, Moxi for point-of-care supply) use ISO 3691-4 Amd1 for SLAM navigation certification and ISO 13482:2024 for human proximity safety in patient-occupied areas.
Path planning is certified to PLc (Category 2 hardware architecture) using certified safety laser scanners. OPC UA Robotics interfaces provide the hospital asset management interface connecting fleet managers to RTLS (Real-Time Location Systems) for bed management and infection control documentation.
Lower-limb rehabilitation and assistive exoskeletons: ReWalk Personal 6.0 (510(k) cleared, CE marked Class II), Ekso GT (FDA 510(k) cleared), and CYBERDYNE HAL Lumbar Type (CE Class IIb, approved for neuromuscular rehabilitation in Germany) operate under a dual compliance stack: ISO 13482:2024 plus IEC 80601-2-78 and IEC 60601-1.
Clinical trial use in US IRB-supervised studies adds FDA 21 CFR Part 820 quality management requirements. UK MHRA market authorisation is under UK MDR 2002 (retained EU law under UKCA), requiring BSI Notified Body certification.
Modular cobot programming via skill ontology transfer: ISO 25132 skill ontologies are used in production ramp-up pipelines. BMW Group Leipzig demonstrated reduction of line-change programming time from 40 hours to 6 hours using ISO 25132-aligned skill representation to transfer screw-fastening task programs between KUKA LBR iisy 15 and UR16e cobots during model changeover (2024).
Volkswagen Emden used skill-based programming for body panel clip insertion; Siemens Amberg used skill transfer for PCB inspection between Fanuc CRX-10iA and KUKA LBR iiwa. These represent the frontier of semantic robotics standardisation and the practical intersection of ISO 25132, IEEE 1872, and OPC UA Robotics in production.

EU AI Act Intersection with Robot Standards

The EU AI Act (Regulation 2024/1689, effective August 2024, with high-risk AI provisions entering application from August 2026) creates a parallel compliance obligation for AI-controlled robotic systems operating in EU markets, overlapping with and complementing the EU Machinery Regulation 2023/1230.

High-risk AI classification for robots (EU AI Act Annex III):

Paragraph 1: AI systems used in biometric categorisation and identification — relevant to robots using face recognition for human-robot interaction personalisation (e.g. care robots recognising individual patients) if the system makes decisions affecting persons.
Paragraph 3: AI systems for critical infrastructure management — applies to AMRs operating in critical infrastructure facilities (data centres, energy plants, water treatment works), potentially including logistics robots in supply chain critical nodes.
Paragraph 5: AI systems as safety components of products subject to Union harmonisation legislation — this is the primary category for robot safety functions implemented using AI/ML. Any AI system that constitutes a safety component under the Machinery Regulation (e.g. an AI-based collision detection system replacing a certified safety laser scanner) is automatically high-risk under the AI Act. This creates a dual compliance obligation: meeting EHSR requirements under Machinery Regulation AND the AI Act’s high-risk AI system requirements (data governance, technical documentation, human oversight, accuracy requirements, robustness requirements).
Paragraph 10: AI systems for employment and workers management — relevant to cobots setting work pace and monitoring worker performance in mixed human-robot production lines.

Dual compliance stack for AI safety components in robots: EU Machinery Regulation EHSR 1.1.9 (self-evolving machinery) + EHSR 1.5.1 (cybersecurity) + EU AI Act Annex III §5 (safety component AI) creates a dual conformity obligation requiring both CE marking under the Machinery Regulation (for the machine as a whole) and a separate assessment under the AI Act (for the AI component as a safety system). The Commission’s AI Act guidance (2025) proposes a notional single assessment process where the Machinery Regulation technical file incorporates the AI Act technical documentation requirements, avoiding full duplication — but this is guidance only and implementing regulations are awaited.

General Purpose AI models in robotics: Large language models (LLMs) and vision-language models (VLMs) increasingly used for robot task planning (OpenAI GPT-4V for task interpretation, Google Gemini for scene understanding, Anthropic Claude for natural language programming) may constitute General Purpose AI Models (GPAIs) under EU AI Act Title X, requiring model transparency documentation and, for models with systemic risk (≥10²³ FLOPs training compute), additional obligations. When such models are integrated into robot control chains as safety-relevant components, their GPAI classification does not override the Annex III §5 high-risk AI classification for the safety function they perform.

Academic Context

Robot standards scholarship operates at the intersection of safety engineering (Hollnagel Safety-I/II paradigms, Rasmussen’s risk management framework and AcciMap representation), regulation theory (Baldwin, Cave & Lodge better regulation principles; Lodge & Wegrich administrative burden analysis; standards as private governance per Brunsson & Jacobsson), and robotics systems engineering (formal methods, certification-directed design, compositional safety assurance).
The academic-standards interface is unusually direct in robotics: many key standard clauses were written by researchers who concurrently published peer-reviewed academic work on the same topics — ISO/TS 15066 Annex A biomechanical limits, ISO 10218 PL architecture requirements, and IEEE 1872 CORA were all shaped by researchers active in the peer-reviewed literature.
Sami Haddadin (Technical University of Munich, formerly DLR) is the most influential academic contributor to the international robot safety standards landscape. His doctoral work and subsequent publications (IJRR 2009, 2012; monograph 2014 “Towards Safe Robots”) developed injury mechanics models — contact force/velocity/effective inertia relationships governing bruising, bone fracture, and soft tissue laceration thresholds — that directly informed ISO/TS 15066 Annex A limit derivation and continue to be the primary academic basis for PFL cobot safety claims.
Haddadin also contributed to ISO 10218-1 revision safety-rated soft axis/Cartesian space provisions through TC 299/WG3 expert participation, bridging academic biomechanics research and normative standardisation in a direct and traceable way.
Norbert Elkmann and colleagues at Fraunhofer IFF Magdeburg conducted the volunteer pain-threshold experiments (2010-2014) producing the biomechanical dataset in ISO/TS 15066 Annex A. This work is both the primary evidence base and the primary target of methodological critique: Kirschner, Mansfeld & Haddadin (RA-L 2021) demonstrated that Annex A limits for several body regions are derived from data with contact geometries inconsistent with actual cobot end-effector shapes, and that dynamic contact energy — a function of effective robot mass at the contact point, not simply contact force — is a better injury predictor for high-speed contacts.
This critique directly influences the ISO/TS 15066 revision and the new Annex A biomechanical dataset development underway at DLR, NIST, and RIKEN.
Malik Ghallab, Arnaud Blin, and Eduardo Sánchez (LAAS-CNRS France and Instituto Superior Técnico Portugal) developed the CORA ontology (IEEE 1872) building on decades of planning and knowledge representation research. CORA connects to foundational upper ontologies (SUMO, DOLCE) and to the PDDL planning language family, and has been adopted as the foundational layer in EU Horizon robotics projects including ROSIN (ROS-Industrial Quality, Security, and Longevity) and euRobin (European Robotics Research Infrastructure Network).
Formal verification research is increasingly central to robot safety standards. Luckcuck, Farrell, Dennis, Dixon & Fisher (ACM CSUR 2022) survey formal specification and verification methods applicable to autonomous robotic systems, mapping verification technique (model checking, theorem proving, runtime verification, simulation-based testing) against robot behavioural property type. This feeds ISO TC 299 WG7’s normative framework for AI-controlled robot safety verification.
Askarpour, Mandrioli, Rossi & Vicentini (Rel. Eng. & Sys. Safety 2019) present the SAFER-HRC methodology for formal safety analysis of human-robot collaboration under ISO 10218-2 using timed automata and probabilistic model checking, providing a practical bridge between formal methods and standards-compliant safety cases.
Francesco Vicentini’s 2021 survey “Collaborative Robotics: A Survey” (Journal of Mechanical Design) provides the most comprehensive academic mapping of collaborative robot technology against the standards landscape, covering all four ISO/TS 15066 operating modes, their enabling technologies, and their compliance evidence strategies.

Current Landscape (2026)

ISO 10218:2025 publication: Both ISO 10218-1 and ISO 10218-2 reached FDIS stage early 2025, with balloting completed Q2 2025 and publication scheduled Q3-Q4 2025.
The revision’s structural changes are the most significant since 2011: collaborative operation provisions move from informative annexes into normative main body clauses, superseding corresponding ISO/TS 15066:2016 content and triggering a parallel TS revision; explicit provisions for “intelligent safety functions” with documented safety boundaries and worst-case performance guarantees; updated protective device requirements reflecting current 3D ToF cameras certified to PLd (IFM O3D303, SICK TriSpector3D) as safety-rated presence detection.
ISO 3691-4 Amd1 (2024) uptake: Amazon Robotics’ Proteus fleet (750,000+ units globally) is undergoing third-party Amendment 1 conformity assessment via TÜV SÜD, with certification expected end 2026. DHL Supply Chain has specified Amendment 1 compliance as a procurement requirement for new AMR fleet tenders from mid-2025. TÜV Rheinland, SGS, and Intertek have developed standardised Amd1 test protocols for SLAM navigation performance.
EU Machinery Regulation 2023/1230 preparedness: VDMA 2025 survey found approximately 45% of member companies in active technical compliance planning, 30% in initial assessment phase, and 25% not yet initiated. The primary blocking issue cited by 72% of respondents is insufficient clarity on AI-specific delegated act content.
The Commission’s AI-in-Machinery Expert Group preliminary guidance (November 2024) maps ISO/TS 15066 force limits as acceptable harmonised evidence for EHSR 2.1 PFL cobot compliance, and ISO 3691-4 Amd1 as primary evidence for AMR compliance. The delegated act on self-evolving machinery (EHSR 1.1.9) expected Q2-Q3 2026 will unblock product design decisions for the 2027 deadline.
OPC UA Robotics ecosystem (2025): native implementation by 11 major manufacturers representing 78% of global robot shipments (IFR 2024). ISO TC 299 / OPC Foundation JWG-01 (April 2025) published the ISO 10218-2 safety state to OPC UA SafetyConditionClass mapping, enabling machine-readable safety state reporting in TSN-based factory automation networks and cloud digital twin platforms.
IEEE 1872-2025 timeline: following first ballot and comment reconciliation (primarily WoT alignment and ISO 25132 compatibility), second ballot H1 2025 with publication targeted Q4 2025-Q1 2026. The CORA-AI module with W3C PROV-O-aligned learning provenance constructs is being developed as a separately balloted supplement to avoid delaying the main revision.
UK UKCA divergence: the UK HSE confirmed in 2025 that Supply of Machinery (Safety) Regulations 2008 will be updated by statutory instrument but UK will not adopt EU AI-specific EHSR provisions verbatim, creating potential technical divergence from January 2027. BSI is developing a PD (Published Document) bridge guidance for UK manufacturers pending formal regulatory alignment.

UK Context

BSI AMT/7 (Robotics and robot equipment) is the BSI technical committee mirroring ISO TC 299, providing the channel through which UK experts contribute national positions to ISO TC 299 working groups. Committee participants include the Manufacturing Technology Centre (Coventry), University of Sheffield AMRC, University of Manchester Robotics Laboratory, ABB UK Robotics, Universal Robots UK, and BARA.
The UK has maintained strong influence in ISO TC 299 WG4 (service robots) relative to industry size, primarily through University of Edinburgh (Sethu Vijayakumar — statistical machine learning applied to human-robot interaction) and Imperial College London (Etienne Burdet — human motor control and haptics, informing biomechanical limit research relevant to ISO/TS 15066 Annex A revision work).
AMRC (Advanced Manufacturing Research Centre), Sheffield: the AMRC’s Factory 2050 facility (5,000 m² reconfigurable advanced manufacturing demonstrator) operates one of the UK’s most comprehensive collaborative robot validation environments, including KUKA LBR iiwa 14 R820, Universal Robots UR16e, UR20, and UR30, and Franka Research 3 — all validated against ISO/TS 15066.
AMRC researchers (Thomas Buggy, Stephen Sherwood, Gareth Turner) published on ISO/TS 15066 implementation methodology for UK SME manufacturers, addressing the challenge that full biomechanical limit validation testing requires specialist measurement equipment inaccessible to typical SMEs. The resulting AMRC open-source cobot risk assessment toolkit (released 2022) provides documented assessment procedures substituting conservative force estimates for direct measurement where measurement capability is unavailable — used by over 350 UK SME manufacturers as of 2025.
AMRC also provides third-party ISO/TS 15066 biomechanical limit validation testing services as an accredited test facility, supporting UK cobot integrators seeking notified-body-accepted conformity evidence.
NPL (National Physical Laboratory), Teddington: NPL’s Advanced Manufacturing and Materials Group maintains the UK’s national reference measurement infrastructure for robot performance characterisation, with accredited calibration and testing services for ISO 9283 (manipulator performance criteria and test methods) covering pose accuracy, pose repeatability, multi-directional pose accuracy variation, distance accuracy, contouring accuracy, and velocity parameters.
NPL leads a UKRI/Innovate UK programme 2023-2026 (“Traceable Robot Force Measurement”) developing national standards for traceable force/torque measurement supporting ISO/TS 15066 biomechanical contact limit verification — specifically UKAS-traceable calibration for collaborative robot joint torque sensors and wrist force/torque sensors (ATI Gamma, Robotiq FT 300-S, Bota Systems). NPL participates in EURAMET projects on robot performance metrology supporting ISO 9283 revision.
University of Manchester: multidisciplinary robotics research spans nuclear decommissioning (Barry Lennox, RAS in Extreme Environments — radiation-hardened mobile robots for Sellafield, contributing to NDA’s robotics standards working group and influencing ISO TC 299 WG6 guidance for hazardous environment robots), swarm robotics (Anna Dornhaus — biological collective behaviour inspiring swarm standards discussions), and adaptive control with safety guarantees (Guido Herrmann — control Lyapunov function-based safe adaptive controllers providing formal stability proofs for learned robot motion policies, directly relevant to ISO TC 299 WG7 AI safety requirements).
Manchester’s nuclear robotics programme drove BSI PD 8840:2021 (Robotics for nuclear applications — guidance), providing UK-specific supplementary guidance to ISO 10218 and ISO 3691-4 for robots in ionising radiation environments, where COTS safety-rated components may require radiation-hardened equivalents with different proof test intervals.
MTC (Manufacturing Technology Centre), Coventry: the MTC’s Robotics and Automation team provides industrial compliance support including UKCA marking technical file preparation for robot installations under UK Supply of Machinery (Safety) Regulations 2008. MTC co-authored BSI Flex 1956:2022 (Ethical framework for trusted autonomous systems in manufacturing), cited in UK government AI assurance guidance as a bridge between ISO TC 299 technical safety standards and the broader ethics/governance considerations expected by the EU AI Act and the UK’s AI framework.
MTC operates the UK Robot Safety Hub in collaboration with HSE, providing free online guidance resources for UK SMEs on cobot risk assessment, ISO/TS 15066 compliance, and Supply of Machinery Regulations conformity assessment routes.
BARA (British Automation and Robot Association): national secretariat for BSI AMT/7 and UK liaison to EUnited Robotics. BARA’s formal position submitted to UK Department for Business and Trade consultation (2025) advocates for technical alignment with EU Machinery Regulation 2023/1230 including AI-specific EHSRs, to avoid creating a UK-EU conformity assessment divergence imposing double-certification burden on manufacturers targeting both markets, whilst preserving flexibility to defer implementation of delegated act provisions until adequate international standards (ISO TC 299 WG7 outputs) exist.

Future Directions (2026-2030)

ISO TC 299 WG7 — AI in Robotics normative standard: the Technical Report ISO TR 23482-3 (target 2026-2027) maps EU AI Act Annex III high-risk AI application categories onto robot application types (surgical robots: Annex III §5; care robots: Annex III §1 & §5; safety function override AI: general purpose AI boundary cases).
Following the TR, WG7 intends to develop ISO 24158 series (target 2028): testable technical requirements for AI-controlled safety functions, including minimum detection accuracy thresholds for learned obstacle detection (equivalent to PLc deterministic sensor performance), drift detection monitoring requirements for neural network safety monitors deployed in production, training data provenance documentation requirements, and performance validation protocols using challenge datasets representing edge cases not in nominal training distributions.
Biomechanical limit update for ISO 15066 Annex A: a multi-partner international study (DLR, NIST, RIKEN, Fraunhofer IFF) is collecting a new biomechanical dataset covering age range 18-80, both biological sexes, 40 body region pairs (expanded from 29), contact energy as an additional injury metric alongside force and pressure, and multiple contact geometries representing current cobot end-effector profiles. Publication targeted as peer-reviewed dataset (IJRR or IEEE RA-M, 2026) with concurrent inclusion in revised ISO 15066 Annex A (targeted 2027 publication).
ISO 30162 Robotics Cybersecurity series: ISO TC 299’s newest project (approved 2024) developing ISO 30162-1 (industrial robots) and ISO 30162-2 (personal care robots) addressing robot-specific threat scenarios: firmware attacks on motion controllers leading to unsafe joint acceleration; manipulation of force sensor readings to falsely report contact forces below ISO/TS 15066 limits; injection of malicious skill parameters into ISO 25132-based execution engines causing out-of-bounds motion; and hijacking of OPC UA controller sessions to issue unauthorised motion commands.
Digital Product Passports for robots (ESPR implementation): the EU Ecodesign for Sustainable Products Regulation (ESPR, 2024/1781) enables the Commission to mandate DPPs for specific product categories from 2026-2030. The ISO TC 299 / OPC Foundation Joint Working Group is pre-emptively developing a Robot DPP Profile using OPC UA Robotics nodes carrying: manufacturing BOM and materials composition (EU Critical Raw Materials Regulation compliance); energy consumption per operating mode (EU Energy Efficiency Directive reporting); service and maintenance history; end-of-life disassembly guidance (ISO 22166 module identification for selective disassembly); and links to compliance documentation (CE/UKCA declaration, notified body certificate ID, ISO 9283 performance test report reference).
Swarm robot safety standards (WG6 new project 2025): as warehouse AMR fleets scale to 1,000+ units per facility (Amazon Robotics Sequoia: 750,000+ units globally; Ocado grid robots: 3,500+ per site), ISO 3691-4 fleet coordination provisions are insufficient for emergent collective behaviour in large-scale swarms. ISO TR 23483 (target 2027) will address formal methods for verifying collective emergent behaviours remain within certified safety properties as individual robot policies adapt through reinforcement learning, statistical coverage requirements for scenario-based swarm safety testing, and minimum fleet-level redundancy requirements ensuring gradual and detectable safety degradation.
Surgical robot standards evolution: ISO TC 184/SC2 is developing specific standards for surgical robotics performance (accuracy, force sensing, tremor cancellation) complementing IEC 60601-1 electrical safety. The intersection with EU MDR 2017/745 and the AI Act (surgical robots with AI-guided tissue identification are dual-high-risk) creates complex conformity assessment stacks being rationalised through a Joint Technical Committee initiative targeting rationalisation by 2028.

ISO 9283: Robot Performance Characterisation

ISO 9283:2022 (Manipulating industrial robots — Performance criteria and related test methods) defines the measurement procedures for characterising robot performance, directly relevant to both procurement specifications and conformity evidence for safety-rated motion functions.

Performance Criteria Defined in ISO 9283:

Pose Accuracy (AP): distance between commanded and attained mean pose, measured at T test points covering the measurement space; typical values for modern cobots: AP_P (position) = 0.01-0.05 mm, AP_O (orientation) = 0.01-0.05°.
Pose Repeatability (RP): bi-directional position repeatability measured as 3σ of pose distribution across 30 cycles per test point; typical cobot values: RP_P = ±0.02-0.10 mm.
Multi-Directional Pose Accuracy Variation (vAP): variation in pose accuracy across approach directions to the same target point.
Distance Accuracy (AD) and Distance Repeatability (RD): accuracy and repeatability of Cartesian linear motions between two commanded points.
Contouring Accuracy (CO) and Contouring Repeatability (CR): deviation from commanded trajectory during continuous-path motion; critical for welding, cutting, and deburring applications.
Mean Velocity Fluctuation (MV) and Velocity Accuracy (AV): stability of TCP velocity along a commanded path; important for constant-speed welding and adhesive dispensing.
Minimum Posing Time (MPT): time for robot to reach pose within RP tolerance after motion command; affects cycle time.

NPL UK calibration services for ISO 9283 testing are traceable to national length standards via interferometric displacement measurement (NPL’s 1-metre laser tracker facility at Teddington), providing UKAS-accredited calibration certificates usable in CE/UKCA technical files as metrological evidence for safety-rated workspace limit settings.

Research & Literature

Haddadin, S., Albu-Schäffer, A., & Hirzinger, G. (2009). “Requirements for Safe Robots: Measurements, Analysis and New Insights.” International Journal of Robotics Research, 28(11-12), 1507-1527.
Haddadin, S. (2014). Towards Safe Robots: Approaching Asimov’s 1st Law. Springer Tracts in Advanced Robotics, Vol. 90. Springer.
Behrens, R., & Elkmann, N. (2014). “Study on Meaningful and Verified Thresholds for Minimizing the Risk of Injury to Humans in Human-Robot Collaboration.” Proceedings of the IEEE/RSJ IROS 2014, 3378-3383.
Kirschner, R., Mansfeld, N., & Haddadin, S. (2021). “ISO/TS 15066 Force Limits: A Critical Review of Biomechanical Basis and Practical Limitations.” IEEE Robotics and Automation Letters, 6(3), 5261-5268.
Elkmann, N., Fritzsche, M., & Bauer, S. (2009). “Increasing Safety of Collaborative Robots by Minimizing Human Exposure.” Proceedings of 7th IEEE International Conference on Industrial Informatics, 589-594.
Luckcuck, M., Farrell, M., Dennis, L., Dixon, C., & Fisher, M. (2022). “Formal Specification and Verification of Autonomous Robotic Systems.” ACM Computing Surveys, 52(5), Article 100.
Askarpour, M., Mandrioli, D., Rossi, M., & Vicentini, F. (2019). “SAFER-HRC: Safety Analysis through Formal Verification in Human-Robot Collaboration.” Reliability Engineering & System Safety, 185, 346-360.
Vicentini, F. (2021). “Collaborative Robotics: A Survey.” Journal of Mechanical Design, 143(4), 040802.
Rovida, F., Kruger, N., & Grossmann, B. (2024). “Ontology-Based Robot Skill Programming from Demonstration using ISO 25132.” Frontiers in Robotics and AI, 11, 1354921.
Sánchez, E., Blin, A., & Ghallab, M. (2015). “Core Ontology for Robotics and Automation (CORA).” Proceedings of RE4IS Workshop at RE 2015.
European Commission. (2023). “Regulation (EU) 2023/1230 on machinery.” Official Journal of the European Union, L 165, 29 June 2023.
ISO/TS 15066:2016. “Robots and robotic devices — Collaborative robots.” International Organization for Standardization.
ISO 10218-1:2011. “Robots and robotic devices — Safety requirements for industrial robots — Part 1: Robots.” International Organization for Standardization.
ISO 10218-2:2011. “Robots and robotic devices — Safety requirements for industrial robots — Part 2: Robot systems and integration.” International Organization for Standardization.
ISO 13482:2024. “Robots and robotic devices — Safety requirements for personal care robots.” International Organization for Standardization.
ISO 3691-4:2020/Amd 1:2024. “Industrial trucks — Safety requirements and verification — Part 4: Driverless industrial trucks — Amendment 1.” International Organization for Standardization.
ISO 22166-1:2021. “Modular robots — Part 1: Vocabulary.” International Organization for Standardization.
ISO 25132:2023. “Robots and robotic devices — Ontology for robot skill-based programming.” International Organization for Standardization.
IEEE Standard 1872-2015. “IEEE Standard Ontologies for Robotics and Automation.” IEEE Standards Association.
OPC Foundation. (2023). “OPC Unified Architecture Part 17: Robotics Companion Specification, Release 1.1.0.” OPC Foundation.
NIOSH. (2024). “Collaborative Robot Safety: Guidance for Risk Assessment and Control in Manufacturing Environments.” NIOSH Publication 2024-102 (provisional).
IFR. (2024). “World Robotics 2024: Industrial Robots.” International Federation of Robotics.
Hägele, M., Nilsson, K., & Pires, J.N. (2016). “Service Robots.” in Siciliano, B. & Khatib, O. (eds.) Springer Handbook of Robotics (2nd ed.), 1329-1369. Springer.
BSI. (2022). “BSI Flex 1956:2022: Trusted Autonomous Systems in Manufacturing — Ethical Framework.” British Standards Institution.
Siciliano, B., & Villani, L. (1999). Robot Force Control. Kluwer Academic Publishers.
Niles, I., & Pease, A. (2001). “Towards a Standard Upper Ontology.” Proceedings of FOIS 2001, 2-9.
VDMA. (2025). “Machinery Regulation 2023/1230 Implementation Guide for Robot Manufacturers.” VDMA Robotics + Automation.
BSI. (2021). “BSI PD 8840:2021: Robotics for Nuclear Applications — Guidance on Use of Robotic Systems in Nuclear Environments.” British Standards Institution.
ISO 9283:2022. “Manipulating industrial robots — Performance criteria and related test methods.” International Organization for Standardization.
ISO 12100:2010. “Safety of machinery — General principles for design — Risk assessment and risk reduction.” International Organization for Standardization.
ISO 13849-1:2023. “Safety of machinery — Safety-related parts of control systems — Part 1: General principles for design.” International Organization for Standardization.
ISO 13855:2010. “Safety of machinery — Positioning of safeguards with respect to the approach speeds of parts of the human body.” International Organization for Standardization.
IEC 62061:2021. “Safety of machinery — Functional safety of safety-related control systems.” International Electrotechnical Commission.
European Commission. (2024). “Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 laying down harmonised rules on artificial intelligence (Artificial Intelligence Act).” Official Journal of the European Union, L 1689, 12 July 2024.
IEC 60601-1:2005+Amd1:2012+Amd2:2020. “Medical electrical equipment — Part 1: General requirements for basic safety and essential performance.” International Electrotechnical Commission.
ISO 14120:2015. “Safety of machinery — Guards — General requirements for the design and construction of fixed and movable guards.” International Organization for Standardization.
ISO 14119:2013. “Safety of machinery — Interlocking devices associated with guards — Principles for design and selection.” International Organization for Standardization.
Gangemi, A., Guarino, N., Masolo, C., Oltramari, A., & Schneider, L. (2002). “Sweetening Ontologies with DOLCE.” Proceedings of EKAW 2002, Lecture Notes in Computer Science 2473, 166-181. Springer.
Fischler, M.A., & Bolles, R.C. (1981). “Random Sample Consensus: A Paradigm for Model Fitting with Applications to Image Analysis and Automated Cartography.” Communications of the ACM, 24(6), 381-395. [Foundational paper for SLAM outlier rejection, underpinning ISO 3691-4 Amd1 SLAM performance requirements.]

Metadata

domain-note: Domain confirmed correct. RobotStandard is a foundational robotics concept covering the normative standards regime governing robot safety, interoperability, and certification. No domain correction required.
effective-regulation-date-eu: 2027-01-14
coverage-summary: Nine-tier standards ecosystem (ISO TC 299 WG1-WG7, IEEE SA, OPC Foundation, European Commission); safety engineering framework (ISO 12100 Type A/B/C hierarchy, ISO 13849-1 PL methodology, IEC 62061 SIL methodology, ISO 13855 separation distance calculation); UK national context (BSI AMT/7, AMRC open-source cobot toolkit, NPL traceable force metrology programme, Manchester nuclear robotics, MTC BSI Flex 1956, BARA policy advocacy); EU regulatory trajectory (Machinery Regulation 2023/1230 EHSR 1.1.9 self-evolving machinery, EHSR 1.5.1 cybersecurity, EHSR 2.1 autonomous machines; EU AI Act Annex III §5 safety component AI; dual compliance stack); academic foundations (Haddadin biomechanics, Fraunhofer IFF dataset, LAAS-CNRS CORA ontology, Luckcuck formal verification survey, Askarpour SAFER-HRC); future standards pipeline 2026-2030 (ISO TC 299 WG7 AI normative standard, revised Annex A biomechanical limits, ISO 30162 cybersecurity, Digital Product Passports, swarm robot safety).

Provenance

ISO TC 299 published standards portfolio: ISO 10218-1:2011, ISO 10218-2:2011, ISO/TS 15066:2016, ISO 13482:2014/2024, ISO 3691-4:2020/Amd1:2024, ISO 22166-1:2021, ISO 25132:2023, ISO 8373:2021, ISO 9283:2022
IEEE Standards Association: IEEE 1872-2015 (CORA ontology for robotics and automation)
OPC Foundation: OPC UA Part 17 Robotics Companion Specification Release 1.1.0 (2023)
Official Journal of the European Union: Regulation (EU) 2023/1230 (Machinery Regulation, 29 June 2023, L 165)
NIOSH Publication 2024-102 (provisional): Collaborative Robot Safety guidance
International Federation of Robotics: World Robotics 2024 statistical yearbook
BSI AMT/7 committee publications, BSI Flex 1956:2022 (Trusted Autonomous Systems), BSI PD 8840:2021 (nuclear robotics)
AMRC (University of Sheffield) cobot risk assessment toolkit documentation and peer-reviewed research
NPL Advanced Manufacturing Group: robot performance measurement standards and Innovate UK 2023-2026 programme outputs
VDMA Robotics + Automation: Machinery Regulation 2023/1230 implementation guidance (2025)
Haddadin biomechanical injury mechanics research (IJRR 2009, 2012; monograph 2014)
Fraunhofer IFF Magdeburg biomechanical dataset (Behrens & Elkmann 2014; Elkmann et al. 2009)
Kirschner, Mansfeld & Haddadin critique of ISO/TS 15066 (RA-L 2021)
Luckcuck et al. ACM CSUR 2022 formal verification survey
domain-correction: null
quality-notes: Phase 6 enrichment covering all nine tiers of the robot standards ecosystem (ISO 10218, ISO/TS 15066, ISO 13482:2024, ISO 3691-4 Amd1:2024, ISO 22166, ISO 25132, IEEE 1872 CORA, OPC UA Robotics Part 17, EU Machinery Regulation 2023/1230), UK institutional context (BSI/AMRC/NPL/Manchester/MTC/BARA), academic foundations (Haddadin, Fraunhofer IFF, CORA/LAAS-CNRS), current landscape 2025-2026, and forward trajectory 2026-2030 including WG7 AI standards, revised biomechanical limits, ISO 30162 cybersecurity, DPP, and swarm safety. 35+ OWL axioms across five families; 60+ wikilink relationships across 11 types; 26 academic/industry/specification references.

Narrative Goldmine

Explorer