An autonomous forklift glides through a crowded fulfillment center, calculating its route to the millimeter and adjusting its speed in milliseconds. It represents the pinnacle of modern industrial intelligence. Yet, if a worker on an upper mezzanine leans out over the aisle, or a fragment of clear stretch wrap dangles from an overhead pallet, the machine is effectively blind. Its primary safety mechanism, a flat two-dimensional laser scanner slicing the air at shin-level, perceives the environment as a single horizontal line. To prevent catastrophic errors, engineers must program massive, conservative safety bubbles around these machines, slowing operations to a crawl and choking factory floor productivity.
The disparity between how smart robots have become and how crudely they perceive physical danger has long been the invisible brake on the automation boom. Oslo-based deeptech startup Sonair has introduced an architectural shift that aims to eliminate this blind spot entirely, unveiling the ADAR One: the world’s first safety-certified 3D ultrasonic sensor designed for human-robot collaboration.
A Shift from Light to Sound
By securing independent functional safety certification from exida —a prominent notified body operating under the European Machinery Directive—, Sonair has moved 3D acoustic sensing from a laboratory experiment to an auditable industrial reality. The ADAR One achieves a Safety Integrity Level 2 (SIL 2) and Performance Level d (PL d) rating, satisfying the demanding requirements of the IEC 61496 standard for electrosensitive protection devices.
Unlike traditional safety infrastructure that relies on 2D LiDAR to map a single optical plane, the ADAR One provides a continuous 180-degree by 180-degree volumetric field of view. It accomplishes this not with lasers or cameras, but through a methodology the company calls Acoustic Detection and Ranging (ADAR). By emitting continuous ultrasonic waves via dense, micro-electromechanical systems (MEMS) transducer arrays and evaluating the returning echoes—a technique inspired by echolocating dolphins—the sensor constructs a real-time spatial map up to four meters away.
“The bottleneck to safe human-robot coexistence isn’t intelligence or speed,” says Knut Sandven, CEO of Sonair. “It’s safe perception; knowing reliably under any condition that a human is nearby. This milestone certification marks the first time a 3D sensor has been independently verified to meet that bar using sound instead of light—a new sensing modality that complements cameras where they fall short.”
Overcoming the Blind Spots of Vision
The choice of ultrasound directly addresses the exact environments where optical systems falter. In heavy manufacturing or high-volume logistics, airborne dust, blinding warehouse glare, and highly reflective or transparent glass surfaces routinely degrade camera and LiDAR performance.
Sound waves remain indifferent to these optical disruptions. The ADAR One serves as a deterministic perceptual backstop, operating independently beneath the higher-level machine vision systems and AI motion stacks. If a primary camera is blinded by a sudden change in lighting, the acoustic layer ensures the machine still recognizes the volume of the space it is entering.
To guarantee the absolute reliability required by international safety regulators, Sonair took the unusual step of engineering the sensor’s embedded system entirely in the Rust programming language. The language, prized in modern software architecture for its strict compile-time memory safety, ensures predictable, real-time data processing without the unpredictable latency spikes common in older architectures. The engineering choice paid off: exida’s validation calculated the sensor’s probability of dangerous failure per hour at less than 1.5 × 10⁻⁷.
Upgrading the Next Generation of Fleets
The commercial implications of a certified 3D safety sensor are already rippling through the sector, offering a massive upgrade path for autonomous mobile robots (AMRs) and industrial platforms. Currently, advanced fleet mainstays like the MiR250 AMR rely on high-precision 2D laser scanners to safely navigate dynamic warehouse floors. The introduction of certified 3D acoustic hardware means platforms in this class could soon operate with significantly tighter safety cushions, moving faster through narrow aisles without compromising human safety.
Early adopters are moving quickly to integrate the hardware into commercial fleets. Among them is beRobox, a leader in plug-and-play industrial palletizing systems, which has entered an agreement to deploy the sensor in high-throughput environments where humans and heavy machinery interact in exceptionally tight quarters. Simultaneously, the sensor is finding traction in public-facing automation: Cleanfix Robotics is utilizing Sonair’s technology in the new generation of its RA660 Navi XL autonomous industrial cleaning machines.
Because the ADAR One maps spaces purely through acoustic feedback without capturing visual images, it offers a distinct privacy advantage. The machines can operate autonomously in corporate offices, retail environments, and hospitals without recording confidential data, saving companies from the complex data-masking and compliance frameworks required by standard video surveillance.
As autonomous vehicles and emerging humanoid robots transition out of isolated industrial cages and into shared human environments, the safety frameworks governing them can no longer afford to view the world in two dimensions. By establishing an audited, sound-based foundation for spatial awareness, Sonair has delivered the missing component required for machines to safely navigate a three-dimensional world.
Explore and Compare Safety Features on Anton Robots
Are you planning to upgrade your facility’s safety layout or expanding your autonomous fleet?
- Compare Platforms: Use our robotics comparison tool to evaluate the payload, speed, and standard sensor arrays of the industry’s leading AMRs.
- Find the Right Fit: Browse the complete technical specifications and request quotes for industrial solutions, from high-performance robotic arms like the Universal Robots UR3e to cutting-edge mobile platforms, directly on Anton Robots.