Sensory Robotics Wins SR-1 Certification for Fenceless Cells, Limits Remain Unclear
Sensory Robotics has announced safety certification for SR-1, its vision-based system for monitoring people around industrial robot arms, although the company did not disclose when certification was completed. The cited cULus 1740 approval and ISO 13849 Performance Level d validation clear the product for commercial deployment in regulated manufacturing environments in the United States and Canada, according to the company. This moves SR-1 beyond laboratory development, but it does not establish that every robot cell can operate without fences: detection range, response time, compatible robot models, installation requirements and operating restrictions remain undisclosed.
SR-1 combines 3D vision, additional sensors and real-time safety software to create virtual protective zones around a robot. Those zones change in response to human movement and machine activity, enabling the safety controller to adjust robot behavior rather than depending solely on a fixed cage and an interlocked access gate.
That control concept is generally known as speed and separation monitoring. An external observer tracks the distance between a person and the robot, then commands the machine to slow or make a safety-rated stop when the protective distance is breached. The practical benefit is not simply removal of steel fencing. It is the ability to make the safeguarded space respond to changing positions, potentially returning some floor area to workers or material flow while preserving separation from a moving arm.
Certification addresses the safety function, not the whole cell
Sensory Robotics says certification testing considered equipment failures, interruptions to safety functions and potential risks to workers. ISO 13849 Performance Level d concerns the dependability of safety-related control functions, while the cited cULus 1740 credential supports access to the U.S. and Canadian markets. Together, these credentials provide a procurement and integration basis that an uncertified laboratory prototype would lack.
They are not, however, a universal approval for any combination of robot, tooling, payload and task. The revised U.S. industrial robot safety standard emphasizes that collaboration is a property of the complete application, not of one robot or sensor by itself. Cell-level risk assessment and validation must still account for the arm, controller, end effector, workpiece, surrounding equipment and the ways people can enter the workspace.
That distinction is especially important for conventional high-speed industrial arms. A camera may locate a worker, but the control system also needs enough time and distance to process that detection, issue a command and bring the moving mechanism to a controlled stop. Robot speed, payload, braking performance, sensor uncertainty and the person’s approach speed all affect the required separation.
A detailed National Institute of Standards and Technology analysis of speed and separation monitoring explains that protective distance depends on sensing and control reaction time, robot stopping time, braking travel and uncertainty in the measured positions of both the person and robot. It also notes that integration choices are consequential: the system must account for the robot’s moving links and attached workpieces, not merely the tool-center point reported by a controller.
Retrofit potential is useful but still unverified
Sensory Robotics says SR-1 can be added to existing industrial robotic production cells. If compatibility proves broad, that retrofit approach could matter more to U.S. manufacturers than a safety system tied to a new robot purchase. Existing arms represent substantial installed capital, and modifying their safeguarding without replacing the manipulator could reduce the mechanical and production disruption associated with redesigning an entire cell.
Retrofit claims nevertheless need robot-specific detail. Older controllers may differ in how quickly they accept stop commands, what motion data they expose and whether their interfaces support the required safety functions. Tooling geometry and payload also change stopping behavior and the volume that must be monitored. Sensory Robotics has not named supported robot models or disclosed how SR-1 handles these integration variables.
The company also says SR-1 has attracted interest from automotive, aerospace, logistics, manufacturing and defense organizations. It has received a $1.25 million Department of Defense Small Business Innovation Research Phase II grant and support through the University of Cincinnati’s innovation ecosystem. No buyer, commercial installation or implementation schedule has been identified, so that interest should not be read as evidence of production service.
Sensory Robotics chief operating officer Mark Gagas has said a conventional safety interruption can require a full shutdown and restart lasting as long as 10 minutes. That figure has not been supported by an independent production study, and actual recovery time will depend on cell architecture and process requirements. Dynamic zones could avoid some unnecessary full stops by reducing speed or holding position, but only if the validated control logic permits those responses for the particular task.
SR-1’s certification is therefore a meaningful transition from development hardware toward deployable safety equipment. The next engineering proof will come from documented installations: supported controllers, measured reaction and stopping performance, environmental limits and validated cell configurations. Fenceless operation is not delivered by vision alone; it emerges from the verified chain connecting sensing, safety logic, robot control, actuation and the mechanical realities of each production cell.
By Jonathan Barrett — Editor for AMI’s future mobility and autonomous systems section, with two decades covering robotics, e-mobility, drone-vehicle convergence, and transport mechanical systems.
