Algorized Edge-AI Human-Machine Interaction Engine

Algorized develops people-sensing edge-AI models, described on its site as a nervous system for Physical AI. The company lists its main office in Etoy, Switzerland, with additional offices in Campbell, California and Tokyo. It addresses the problem of machines that must stop or slow down near workers because they cannot perceive people reliably.

The Edge-AI Human-Machine Interaction Engine, unveiled at CES 2026, reads wireless signals from Wi-Fi, UWB and mmWave sensors instead of cameras and runs inference at the edge with low latency. The software works with low-cost off-the-shelf sensors and targets industrial safety, robotics, in-cabin automotive sensing and infrastructure. Partners listed include Qorvo, Texas Instruments and ASUS.

• Multi-modal sensing: models for Wi-Fi, UWB and mmWave signals.
• Edge inference: low-latency processing directly on edge computing platforms.
• Predictive safety: anticipates human movement so machines adapt speed instead of stopping.
• UWB radar: detection through obstacles using the ALGORIZED UWB Radar.
• Broad use cases: industrial safety, robot interaction, in-cabin automotive sensing and infrastructure.

Robot cells in electronics assembly plants where workers and robots share space. Automated guided vehicles and forklifts in warehouses that must detect people behind obstacles. In-cabin occupancy sensing for vehicle makers.

Detects people without cameras, which avoids video privacy concerns. Allows robots to keep running at full speed near workers, reducing costly stops. Software-only approach runs on low-cost off-the-shelf sensors. Low-latency inference at the edge.

Collaborative robots and automated lines in factories need reliable human detection to meet safety requirements without losing throughput. Vietnamese and Southeast Asian electronics and automotive plants that are adding robots face the same trade-off between safety stops and productivity.

Headquarters in Etoy, Switzerland; offices in Campbell, CA and Tokyo. $13M Series A reported 2026-02 (Run Ventures lead, Amazon Industrial Innovation Fund, Acrobator Ventures). Berkeley SkyDeck cohort and TechCrunch Startup Battlefield 200.