We research and develop intelligent biosensors, adaptive robotic kinematics, and 3D geometry agents to redefine product development and mechatronic control processes.
Our active deep-tech projects cover mechatronic innovations, biosignal processing, and generative spatial AI models.
Combination of multimodal LLMs and spatially trained GNNs for automated generation and FEM optimization of complex 3D geometries based on text instructions.
A wearable sensor armband that translates muscle activity signals (sEMG) into precise navigation and editing commands for CAD modeling in real-time using edge classifiers.
An industrial 3D printer with a 500mm build volume that automatically folds down to suitcase size using kinematic scissor mechanisms. Constant belt tension without recalibration.
A mobile camera application to assist visually impaired individuals in recognizing physical game elements. Local object detection and low-latency voice output.
All of our system components, from biosensors to kinematics, are indexed in a high-dimensional vector database. The system determines functional similarities via geometric cosine distances.
Set a **search vector (Q)** by clicking in the coordinate system, or use our presets to map the calculated similarity to the technological clusters.
We are currently evaluating pilot applications for the EMG-CAD neuro-interface and licensing models for the CAD-AI agent system. Register for regular technology updates.