3D ultrasonic sensor earns SIL2 PL d safety approval

A tiny 3D sensor just earned robots a safety certification.

Sonair’s ADAR One is now certified for SIL2 under IEC 61508 and Performance Level d under ISO 13849, with compliance to IEC 61496 for electrosensitive protection devices and the European Machinery Directive. The sensor is designed to be embedded flush into virtually any robot form factor, enabling safer navigation around people and objects in high‑risk industrial environments. The certification marks a rare convergence of AI capability and formal safety validation, a pairing executives have long sought but rarely achieved at scale.

The case study reports that the safety certification process was extensive, demanding a rigorous trail of tests and documentation. Knut Sandven, CEO of Sonair, described the effort as a multi‑month, cross‑discipline sprint that required pausing other development to satisfy every requirement. In Germany, exida served as the assessing body to verify the system’s alignment with SIL2 and PL d targets. Deployment data shows the industry is beginning to demand safety‑first perception stacks that go beyond traditional 2D perimeter sensing, and the ADAR One stands as a blueprint for how 3D sensing can be integrated without forcing a redesign of existing robot platforms.

The core advantage of the ADAR One is its 3D awareness, addressing a long‑standing gap in robot safety. Traditional 2D safety scanners define a plane, then project risk management onto that single surface. As Sandven notes, “Navigation in 2D can be perfectly fine, but safety in 2D is a different matter.” A robot relying solely on a flat safety perimeter can miss hazards above or below the scanner plane, or around occlusions at different heights. By contrast, the 3D ultrasonic approach detects humans and objects across vertical dimensions, improving real‑world safety when people move dynamically in workshop aisles, around loading docks, or near heavy equipment.

For plant managers and CFOs evaluating automation investments, the ADAR One signals a refocusing of risk and reliability as a first‑order ROI driver. The case study underscores that while the certification does not automatically translate into speed or throughput gains, it reduces the envelope of risk that constrains autonomous operations in human environments. In practice, deployments will hinge on how quickly the sensor can be integrated with existing control systems and how it interoperates with robot safety controllers, emergency stops, and machine‑vision workflows. One clear integration note is the need to embed the sensor flush into a robot’s chassis, ensuring robust protective housing and minimal intrusion into the robot’s geometry and field of view.

Two important practitioner insights emerge. First, the integration requirements are nontrivial: a 3D perception sensor must be matched with a compatible safety controller, power budgeting, and software that can fuse ultrasonic data with existing LiDAR or vision streams without compromising real‑time performance. Second, the cost of certification and validation must be weighed against the value of safer operations and uptime. As an evidence check, deployment data shows the trend toward more capable human‑detection sensors is accelerating, but the case study stops short of publishing cycle times or throughput figures for ADAR‑equipped systems. That leaves a cautious but optimistic path for operators who are willing to invest in end‑to‑end safety architecture rather than a quick, one‑off sensor swap.

What to watch next: the emergence of standardized integration patterns, shared safety profiles across robot families, and case studies that quantify safety‑driven reductions in downtime and incident risk. If the industry keeps pushing toward 3D sensing with formal safety validation, expect more mobile robots to operate closer to people, with predictable behavior and measurable gains in real uptime.