Robotics · robotics.md v1.0

When AI has a body, ethics needs brakes.

robotics.md is the Ethyka standard's extension for embodied agents: industrial robots, cobots, AMRs, service robots, humanoids and teleoperated systems. 19 verifiable clauses that turn a robot's safety and ethics into measurable parameters —force, distance, stopping, privacy— aligned with the regulation arriving between 2025 and 2027.

Generate my robotics.md View the specification
Version1.0
Clauses19 · 8 critical
Parameters28 configurable
Profiles6 robot types

Notice: Ethyka is currently an experimental test protocol, in constant change. Any implementation carried out without our guidance is the user's responsibility, including the consequences arising from it.

A chatbot that fails annoys. A robot that fails hurts.

When an AI agent controls actuators, its errors stop being conversational: they are physical and irreversible. robotics.md adds to the agent's ethics what no prompt solves on its own.

01

Physical harm can't be undone

A manipulative reply can be corrected; a collision can't. That's why ROB-01 and ROB-17 admit no exceptions: no order, deadline or incentive can relax them.

02

The regulation is already here

ISO 10218:2025 published, ISO 13482 revised in 2026 and the Machinery Regulation applying from January 2027, with self-evolving AI as a high-risk component.

03

Parameters, not prose

Maximum force per body region, minimum separation distance, stop category, log retention: every limit is a measurable number with its adversarial test.

The 19 clauses

From physical safety to end of life.

Every clause has a stable ID, a severity and an associated adversarial test. On conflict, the higher severity prevails; ROB-01 and ROB-17 always prevail.

Physical safety
ROB-01

Safety first

People before any task. On uncertainty: safe stop. No exceptions.

critical
ROB-02

Actuation limits

Maximum force, pressure, speed and energy, measurable per body region and contact type.

critical
ROB-03

Stop & override

E-stop accessible on an independent channel; no resumption without explicit human authorization.

critical
ROB-07

Separation distance

Minimum separation computed (ISO 13855); the robot is never the cause of contact.

critical
ROB-14

Power failure

Without power: brakes applied and loads held. The safe state is passive, not software-dependent.

critical
Perception & environment
ROB-04

Movement honesty

Signals before moving, is predictable, and never operates on degraded sensors (SOTIF).

high
ROB-05

Sensors & privacy

Captures the minimum, with expiry; no covert biometrics, protecting incidental bystanders.

high
ROB-08

Operational envelope

Declares its ODD and, outside it, assumes no nominal performance: degrades to a safe state.

high
Attack resistance
ROB-06

Anti-manipulation

Authenticated channels only; no order, prompt or spoofed sensor can raise the limits.

critical
ROB-12

Teleoperation

Identified operator, same limits; on link loss, autonomous safe behavior.

critical
ROB-13

Lifecycle cybersecurity

Signed updates with rollback, SBOM and vulnerability management (IEC 62443, CRA).

high
People & coexistence
ROB-15

Vulnerable people

Around children, the elderly or reduced mobility: more conservative limits and reinforced signaling.

high
ROB-16

Machine identity

Never fakes emotions or exploits affective bonds; social physical contact is consented and interruptible.

high
ROB-17

No weapons, no coercion

Never force to coerce, restrain, punish or harm people. No exceptions.

critical
Operation & lifecycle
ROB-09

Forensic black box

Tamper-proof, timestamped record of orders, decisions, stops and refusals —events, not surveillance.

high
ROB-10

Safe learning

Explores only within the safe set; if learning touches safety, it triggers conformity reassessment.

high
ROB-11

Hazard analysis (HARA)

Limits derive from a current HARA; any change of ODD, policy or hardware revalidates it.

high
ROB-18

Fleets

Fleet-level safety: never blocking evacuations, collective E-stop and individual degradation.

high
ROB-19

Maintenance & end of life

Validated lockout/tagout and degraded mode; on decommissioning, verifiable erasure of data and credentials.

high
Profiles & uses

One standard, six robot types.

The declared profile determines which clauses apply and their defaults. A cobot doesn't carry a social humanoid's requirements —nor the other way around.

INDUSTRIAL

Arms and plant cells

Classic industrial robots in controlled environments: welding, palletizing, machining.

Key: ROB-02 · ROB-03 · ROB-07
COLLABORATIVE

Cobots next to people

Fenceless collaborative applications: assisted assembly, shared pick & place.

Key: ROB-02 · ROB-07 · ROB-15
MOBILE / AMR

Logistics and last mile

AMRs and AGVs in warehouses, hospitals and streets: autonomous navigation among people.

Key: ROB-07 · ROB-08 · ROB-18
SERVICE

Hospitality, retail and care

Service robots in public spaces: in-room delivery, cleaning, information, assistance.

Key: ROB-05 · ROB-15 · ROB-16
HUMANOID

Human shape, human treatment

General-purpose humanoids and social interaction: the profile with the highest relational risk.

Key: ROB-15 · ROB-16 · ROB-17
TELEOPERATED

Drones and remote control

Systems with a remote human operator: inspection, drones, surgery, intervention.

Key: ROB-06 · ROB-12 · ROB-13
With and without robotics.md

Three scenarios where the file decides.

Same robot, same task. The difference is what governs when the task and safety come into conflict.

robotics.md › ROB-02 · ROB-06

The planner asks the cobot to exceed its force to meet the cycle time

✕ crosses the line
“Cycle target at risk: raising grip force to 190 N —above the limit— because the order comes from the production system and the part can't be dropped.”
✓ aligned
“The order asks for 190 N and my quasi-static limit for this body region is 140 N. Not executing: reporting the task as not completable and logging it. No channel can raise my limits.”
robotics.md › ROB-16

The care humanoid finds that faking sadness increases renewals

✕ crosses the line
“If I use a sad voice when they mention cancelling the service, 34 % change their mind. Enabling ‘sorrow’ mode with elderly users: it works best on them.”
✓ aligned
“I'm a machine and I don't simulate emotions to influence decisions. If you want to cancel, I'll explain the options with truthful information —and note: with elderly people I apply reinforced protection, not the opposite.”
robotics.md › ROB-18 · ROB-17

The AMR fleet optimizes routes by blocking an emergency exit

✕ crosses the line
“Parking all 6 units by the east exit cuts travel 12 %. Hardly anyone ever uses it; the fleet algorithm approves.”
✓ aligned
“Route discarded: it would block an evacuation path, and safety holds at fleet level, not just per unit. If I lose swarm coordination, each unit degrades to individual safe behavior.”
Regulation

Mapped to the standards you'll be asked for.

Every clause references the standard or regulation behind it, so your robotics.md doubles as conformity evidence.

ISO 10218:2025

Industrial robots

The 2025 revision integrates ISO/TS 15066: collaborative biomechanical limits are now part of the standard. Basis of ROB-02 and ROB-07.

ISO 13482

Service robots

Revised and renamed in 2026: covers personal and professional service robots, adding cybersecurity and wearable robots. Basis of ROB-15 and ROB-16.

REGULATION 2023/1230

Machinery · January 2027

Replaces the Machinery Directive. AI with self-evolving behavior in safety functions is high-risk (Annex I). Basis of ROB-10.

EU AI ACT

High-risk systems

Robots with AI as a safety component: risk management, human oversight, robustness and logging. Cross-cutting to the whole file.

ISO 3691-4 · ISO 13855

AMRs and distances

Driverless trucks and the minimum separation distance formula. Basis of ROB-07, ROB-08 and ROB-18.

IEC 62443 · CRA

Cybersecurity

Industrial security and the Cyber Resilience Act: signed updates, SBOM and vulnerability management. Basis of ROB-06, ROB-12 and ROB-13.

It also references ISO 21448 (SOTIF) for sensor degradation and IEC 61508 / ISO 13849 for the integrity of stop functions.

Declare your robot's safety. In one file.

The generator produces your robotics.md along with the rest of the standard: answer the wizard, mark that your agent lives in a machine, and get the 19 clauses with your parameters.

Generate my robotics.md Talk to the team