We turn selected research opportunities into inventions, intellectual property, and commercially relevant technologies. Each program follows its own path.
Distributed & Hybrid Computing
Current technology · Edge & Hybrid AIoT
Architectures that let AI systems operate under limited connectivity and coordinate local processing with cloud intelligence.
Potential applications: edge AI devices, autonomous systems, and connected products.
Multimodal Sensing & Fusion
Approaches to coordinating connected devices and combining multiple sensor signals into coherent, context-aware behavior.
Potential applications: intelligent environments and context-aware physical systems.
AIoT & Intelligent Hardware
Hardware architectures that integrate processing chips, sensors, memory, and AIoT components into one coordinated system.
Potential applications: intelligent devices, embedded AI, and connected hardware products.
Biosensing & Biological Systems
Biochemical sensing combined with AI methods to interpret measurement signals.
Potential applications: research instrumentation and analytical systems, subject to validation.
Technology candidate families under evaluation. Not completed inventions, granted patents, or validated products.
One of our three core technologies: human-governed AI agents, built on the belief that humans and agents work as one system. Our research addresses two central problems in trustworthy autonomy, and is being applied, together with Human Agentic Systems, through LUP Systems.
Contextual Integrity
Retrieving exactly the context a specific outcome needs, without compromising data boundaries.
Governed Autonomy
Service-oriented architectures where every autonomous action is authorized, traceable, and interruptible by a human.
An R&D program investigating agent execution, data governance, human authority, and verifiable controls across heterogeneous computing environments.
We invent technologies where autonomous agents act within clear human authority.
How can agents operate within clear boundaries and human approval?
How can agent actions be authorized, traced, evaluated, and safely interrupted?
How can systems preserve reliable context, evidence, and decision history?
How can agents combine voice, vision, sensor data, and computing resources across environments?
How can research become reusable, protectable, and commercially viable technology?
We welcome researchers, engineers, and universities for joint R&D, technical collaboration, and invention development.
We approach invention as a continuous learning process: understand the problem, test ideas, build evidence, consider appropriate protection, and explore practical routes to use.
Identify meaningful problems and opportunities.
Turn concepts into testable technical work.
Choose suitable IP and confidentiality measures.
Engineer and validate integrated systems.
Find the right route to value.
Each program is designed to add know-how and evidence, along with protectable IP where appropriate, that can strengthen what we build next.
License or acquire suitable technology and IP under agreed terms.
Turn validated technology into integrated products.
Advance shared research questions with clear scope and agreed IP terms.
Develop and scale technology together with partners.