uMe — humanoid robotics that delivers what it promises

What uMe can do – today

uMe is a humanoid platform, not a concept study. Everything on this page is either available now or clearly marked as roadmap. We do not promise what the robot cannot do.

• Voice interaction – natural conversation via microphone array and speaker, in German and English, extendable to further languages.


• Display communication – content, prompts and visual support on the integrated screen.


• Teleoperation and guided use – safe operation with a human in control.


• Prepared skills – learning modules, dialogues and movement sequences, configurable through uGo+ without programming.


• AI-based perception – object, landmark and human detection via depth camera, processed locally on the device.

What stays on the device

Perception runs locally. Camera image, object and person recognition are processed on the robot itself – with 200 TOPS of computing power on board. This data does not leave your building.

The voice dialogue currently uses an external AI service. What is spoken is processed there. We say so openly, because it is the decisive point in your data protection review. A fully local voice dialogue is in development. You decide which functions are active: uMe can be set up so that the voice dialogue is switched off and only the local functions run.

Status at a glance

The LED ring on the head shows the battery state without anyone having to read a display or open an app – visible from ten metres:

• Normal operation – enough charge for the task at hand.


• Charge soon – the ring changes as the battery runs low. Time to plan the next charge.


• Critical – the ring signals clearly that uMe needs to return to the charging station.

The logic is identical across all three variants.

Three variants, one system

All variants share the same platform, the same software and the same safety concept. They differ in scope – and can be extended step by step.

• uMe Uno – the base platform for education, research and demonstration. Voice, display, teleoperation, prepared skills, uGo+ readiness.


• uMe Duo – Uno plus gripper package. For first manipulation, training scenarios and hand-over interaction.


• uMe Trio – Duo plus LiDAR navigation package. Mapping, localisation and guidance within pilot scope.

Where uMe is used

• Education and training – robotics, AI and automation on real hardware instead of in simulation.


• Universities and research – open interfaces, sensor data, logging and reusable workflows for HRI studies and custom applications.


• Hospitals and clinics – reception, wayfinding, recurring information and reproducible training scenarios.


• Care and assisted living – structured conversation, activation, reminders and social presence.

uMe does not replace care staff and does not make medical decisions. Suitable use covers information, orientation, training and non-clinical routines.

An honest roadmap

Available today: voice interaction, display communication, teleoperation, prepared skills.

Advancing next: LiDAR navigation, sensor fusion, collision avoidance, semantic understanding.

Building further: tool changer, assistance tasks, advanced workflows, partner ecosystem.

Technical data

• 230 × 470 × 1,200 mm, from 30 kg depending on configuration


• 23 degrees of freedom, omnidirectional walking, fall detection and get-up routines


• NVIDIA AGX Orin with up to 200 TOPS, 14-core CPU; perception is processed on the device


• Depth camera, 9-axis IMU, circular 6-microphone array, integrated speaker


• Up to approx. 6 hours of operation, 48 V charging station


• Wi-Fi 6, Bluetooth 5.2, optional 5G; REST, MQTT and WebSocket interfaces, ROS2 compatible


• IP43 (battery IP66), CE according to Machinery Directive 2006/42/EC and ISO 13482

Availability

uMe can be ordered now. Delivery within approximately 8 weeks. Configuration, integration scope and service are agreed in an initial conversation.