Risk assessment is the systematic process of identifying what could go wrong with a robot and reducing each hazard to an acceptable level — the required first step before any robot is deployed near people.
Risk assessment is thinking through every way a robot could hurt someone or cause damage, then adding safeguards until each risk is acceptably low. It's the required homework before a robot is allowed to work around people.
🎯 Quick challenge
The preferred first choice in reducing a robot hazard is to…
Before any robot is allowed to work near people, one process comes first — and it's required by law and standards worldwide: risk assessment. It's the disciplined way of asking "what could go wrong, and have we made it safe enough?"
The process
Risk assessment (formalized in ISO 12100 for machinery and ISO 10218 for robots) follows a systematic loop:
Identify hazards. List every way the robot could cause harm — crushing, impact, pinch points, sharp tools, electrical, the workpiece, unexpected motion, and interactions with people.
Estimate the risk of each — combining severity (how bad the harm) with probability (how likely, how often exposed, how avoidable).
Reduce the risk with safeguards, then re-assess the residual risk. Repeat until every risk is acceptably low.
Identify, estimate, reduce, repeat
The cycle continues until residual risk is acceptable. It's iterative — each safeguard changes the picture and is re-evaluated.
The hierarchy of controls
Not all safeguards are equal. Risk reduction follows a strict priority order:
Inherently safe design first — eliminate or reduce the hazard itself (remove pinch points, limit force and speed as in power and force limiting, round edges). Most reliable, because it can't be bypassed.
Information for use — warnings, training, signage, procedures. Weakest, because it relies on people behaving correctly.
You always prefer the highest level that's feasible — a warning label is a last resort, not a substitute for a guard.
Why it's non-negotiable
It's required. Regulations and standards mandate a documented risk assessment before deployment; it's central to CE marking and legal compliance.
It's specific. The same robot in two different cells has different risks — the assessment is about this application, not the robot in the abstract.
It drives the design. It determines which safeguards (SRMS, fencing, force limits) are needed and to what safety integrity level.
Why it matters
Risk assessment is the foundation of all robot safety — the structured discipline that turns "seems fine" into "demonstrably safe enough." Every safeguard on a robot exists because a risk assessment called for it. It's the essential, mandatory groundwork for deploying robots responsibly around humans.
