Genisom AI ships 10,000 deployable robot platforms in under three years

Genisom AI shipped more than 10,000 deployable robot platforms in under three years, a milestone it showcased at ICRA 2026 in Vienna. The Beijing-based startup has moved from concept demos to real world deployments with mass produced systems, a rarity in the robotics space where many players struggle to scale beyond lab tests. Since its founding in December 2023, Genisom says it has delivered a production scale library of platforms, aimed at bridging hardware and software into deployable industrial solutions rather than one off demonstrations.

The company presented two core platforms at the conference, the M1 quadruped and its L1 series, portraying them as mature, field ready systems. The M1 is pitched as an industrial grade robot with a continuous walking payload of 30 kilograms and a payload to weight ratio near 1 to 1. The platform carries an IP67 protection rating, is capable of up to five hours of runtime depending on payload and operating conditions, and relies on Genisom AI’s in house P85MAX-S joint actuator module. That actuator module can deliver up to 180 newton meters of peak torque within an 86 millimeter form factor that weighs roughly one kilogram, giving the company tighter control over the hardware software integration points that influence locomotion reliability and payload performance. Genisom says in house actuation supports more predictable behavior and faster iteration cycles when integrating perception, planning, and control into a working system.

Manufacturing for real world deployment is a core claim for Genisom AI. The company’s emphasis on platform based design mirrors peers like Unitree, but Genisom stresses its own in house core technologies and real world application capabilities as a differentiator. At ICRA, the M1 and the L1 series were presented as mature, mass produced systems, underscoring Genisom’s commitment to moving beyond laboratory demonstrations toward scalable customer delivery. The M1 Ultra, the company notes, adds an advanced capability beyond the baseline M1, though details were not fully disclosed in the show materials.

From an operations perspective, the numbers behind Genisom AI's growth matter for readers weighing automation investments. Deployment data shows the company has reached a production cadence that supports a substantial installed base in a short time, a factor CFOs and plant managers watch closely when calculating return on investment. The sheer scale, more than 10,000 units, helps drive unit economics down and shorten payback periods, especially for tasks that align with a quadruped's versatility in uneven terrain or cluttered environments where wheels or fixed arms struggle. The question for asset owners remains how these platforms perform on specific tasks, how quickly they can be re tasked, and how integration with existing lines and safety protocols is managed.

Practitioner insights:

The ICRA reveal positions Genisom AI as a late stage producer with a clear eye on real world deployment, not a carousel of lab demonstrations. The company’s stated path, industrial platforms with in house hardware and software integration, designed for scalable deployment, addresses a central concern for plant executives: can automation scale without eroding reliability or inflating integration risk? The proof, for now, rests in the numbers, a growing installed base and a track record of delivering tens of thousands of units in a little over two years. What remains to be seen is the cycle time performance for specific applications and how quickly users can convert platform capabilities into measurable throughput improvements on the line or in the field.