Mantis MR-1 Safety Certification: What It Proves

Mantis Robotics announced on February 24, 2026 that its MR-1 robot had received certification to ISO 10218-1 and ISO 13849-1. The company also states that the robot carries CE marking. For integrators, this is a product milestone, not automatic approval for every fenceless application.

What the Announcement Confirms

The manufacturer's announcement identifies the MR-1 as a high-speed industrial robot designed for fenceless operation. It states that the product was certified to the two named standards. Mantis also describes embedded safety functions and environment sensing.

Those statements concern the robot and its safety-related controls. They do not replace a complete machine risk assessment. Tooling, workpieces, fixtures, process hazards and surrounding equipment remain part of the integrated application.

Robot Certification and Cell Safety Are Different

ISO 10218-1 addresses safety requirements for industrial robots. ISO 13849-1 provides methods for designing safety-related control-system parts. A certified robot can support a compliant cell, but integration decisions determine the final risk.

Sharp tools, hot parts, heavy workpieces and crushing points can make close human access unacceptable. Material ejection can also require physical containment. A fenceless layout is therefore an engineering outcome, not a product feature that applies everywhere.

The integrator must identify tasks across the full lifecycle. Programming, recovery, cleaning, maintenance and tool changes can expose different hazards from automatic production.

Speed Changes the Safety Calculation

Mantis promotes the MR-1 as capable of industrial speed while sensing nearby people. Higher speed can improve cycle time, but it also increases stopping distance and transferred energy. Payload, tool mass, robot posture and braking behavior all affect the result.

Safety validation must use the permitted speed for each operating condition. Tests should include the worst credible payload and extension. A demonstration with an empty wrist does not prove performance with production tooling.

Protective separation depends on detection range, response time and stopping performance. Floor layout and approach direction also matter. Reflective surfaces, occlusion and environmental changes should be tested under real cell conditions.

Understand the Safety Functions

A long list of safety functions is useful only when each function has a defined role. Engineers should map hazards to required functions, performance targets, inputs, logic and final switching elements.

Typical functions may limit speed, position or motion. They may also request a protective stop. The exact MR-1 functions, limits and configuration rules must be verified in current manufacturer documentation.

Diagnostic coverage does not eliminate proof testing. Fault detection, reset behavior and recovery must be validated. A reset should not initiate hazardous motion or bypass a failed sensing channel.

Perception Needs Defined Boundaries

Environment sensing can help a robot respond to changing human position. It should not be described as human-like judgment. The safety case must depend on specified, testable behavior rather than broad artificial-intelligence language.

Document the detection zone, minimum object characteristics, occlusion limits and response to sensor degradation. Test clothing, carts, pallets and tools that appear in production. Lighting, dust and background geometry may affect non-contact sensing.

If a safety function relies on a perception channel, its fault response must be clear. The machine needs a defined condition when confidence is lost or a sensor becomes unavailable.

Tooling Can Dominate the Risk

A low robot payload does not guarantee low application risk. End effectors can contain pinch points, vacuum loss hazards, stored pneumatic energy or sharp edges. The carried part can be larger than the robot envelope.

Use passive design measures before relying on control functions. Rounded tooling, retained fasteners and limited stored energy can reduce consequences. Guards may still be required around process-specific hazards while other cell areas remain open.

Integration With PLC and Line Controls

Define ownership of start, stop, reset and mode selection. Safety signals should use the architecture and interfaces specified by the manufacturer. Standard network status must not replace a safety-rated signal.

Coordinate upstream conveyors, fixtures and downstream machines. Stopping only the robot may leave another hazard active. Recovery sequences should identify which device can move and who controls restart.

Related components can be reviewed in the Industrial Detectors and ABB Automation collections. These links cover related sensing and robotics hardware, not equivalent products.

Commissioning Evidence Matters

Preserve the approved safety configuration and software versions. Record the tool, payload, stopping-test method and results. Validate every operating mode and every authorized access route.

Challenge detection zones at representative speeds and positions. Test power loss, communication loss and sensor faults. Confirm that restart requires intended conditions and that diagnostics guide recovery without encouraging bypasses.

Changes require review. A heavier gripper, faster program, moved workstation or new pallet can invalidate assumptions. Configuration management should connect drawings, risk assessment, safety settings and validation records.

Procurement and Lifecycle Questions

Buyers should request current declarations, certificates, manuals and supported software versions. Integrators should confirm training, spare strategy and support arrangements before committing a production line.

Cybersecurity also belongs in the lifecycle plan. Record enabled services, access roles and backup methods. Firmware changes should follow controlled testing because they can affect diagnostics, interfaces or validated behavior.

What Engineers Should Take From the News

The certification gives integrators stronger product-level evidence for evaluating the MR-1. It does not make cages obsolete. The useful change is the possibility of more open cells when measured risk and validated functions support that decision.

Mantis lists the announcement date and certification scope in its official February 2026 release. The company provides current product information on the MR-1 page. Teams should use the latest manuals and certificates before specifying a production cell.