Situational and self-awareness technologies

We conduct advanced research on technologies that identify, interrogate, and evaluate both environmental conditions and system states, enabling AI agents to make intelligent, data-driven decisions.

The challenges we tackle

1

Realising autonomous systems and robots

We are working on intelligent perception systems that enable autonomous systems to make smart decisions, become more flexible and navigate complex, dynamic environments, and that enable industrial robots to understand their environment and adapt their behaviour accordingly.

2

Automated visual quality inspection

We automate visual quality control with computer vision that detects, classifies and localises defects in real time, and work on tools for the rapid development of smart AI agents that accelerate this automation.

3

World modelling for autonomous systems

Autonomous systems only take the right actions when they truly understand their driving, sailing or flying environment. We work on solutions to continuously build semantic world models that bring together and interpret various sensor data.

4

Condition monitoring and anomaly detection

We work on systems that run with less and less supervision. AI agents monitor their condition, detect deviations, find causes, and predict and plan maintenance and repairs.

We focus on

Sensing and perception

Sensing and perception technologies collect and process information from sensors and instruments in and around a system. This creates situational awareness for robots, autonomous systems and industrial processes. The next generation of perception systems works together with AI agents, which also add semantic interpretation. This enables machines to truly understand their environment and respond intelligently.

State estimation

‘Status’ encompasses everything that matters for a system: position, orientation, speeds per degree of freedom, health, and maintenance requirements. Estimating internal and external states from multi-sensor data, tracking their evolution, and predicting future values is crucial for intelligent autonomy, as this determines how systems behave and interact with their environment.

Knowledge & Model Building

Knowledge and modelling technologies create digital representations of systems, processes and their environment by integrating all available information into a single central model.

These representations, built from sensing, perception and human interaction, can be consulted to support intelligent decision-making and autonomous behaviour.

Hazard Assessment

Autonomy requires safety first. Risk assessment technology evaluates whether the environment, the system, or their interactions create risks that could undermine planned (in)action, the system, or the mission. Safety must be embedded in the architecture, guiding every decision and monitoring every interaction.

Anomaly detection

You cannot predict everything in advance; anomalies will occur. Detecting them quickly is essential for the safety and efficiency of autonomous and production systems.

Anomaly detection is the crucial first step in risk mitigation and an important key to greater system intelligence and autonomy.