What we do

Multi-physics, data-driven and hybrid models of electrical, thermal, mechanical and fluid systems — the basis for design decisions, virtual testing and digital twins.

• Physics-based and grey-box models
• System identification from test data
• Reduced-order and real-time capable models
• Digital twins for virtual testing

Neural-network and probabilistic models for prediction, anomaly detection and computer vision — on time series, sensor signals, images and RF, deployed down to the edge.

• Prediction, anomaly detection and classification
• Computer vision and signal processing
• Hybrid physics + ML models
• Real-time inference on embedded controllers
• Edge inference on embedded GPU / edge AI platforms
• MLOps: training, deployment and monitoring

Model predictive, robust and adaptive control for powertrains, energy and thermal systems, actuators, rotating machinery and precision mechatronics.

• Model predictive control (MPC)
• Robust and adaptive control
• Disturbance rejection and constraint handling
• Real-time implementation-aware design

State and parameter estimation, virtual sensors, navigation and localization — robust to noise, outliers and missing data.

• Optimal state estimation and Kalman filtering
• Virtual sensors
• Multi-sensor fusion, navigation and localization
• Tracking for unmanned and autonomous systems

Model-based and machine-learning fault detection, prognostics and remaining-useful-life estimation.

• Fault detection and isolation
• Degradation modelling and prognostics
• Remaining-useful-life estimation
• On-board and cloud-based monitoring

Data pipelines from test benches and fleets to calibrated models, dashboards and decisions.

• Test-bench and fleet data pipelines
• Model calibration and recalibration
• Dashboards and engineering analytics
• Optimization under performance, energy and lifetime constraints

Embedded implementation on target hardware, the measurement chains that feed it, and functional verification & testing.

• Code generation and C/C++ implementation
• Integration on real-time platforms
• Instrumentation and measurement chains: sensors, signal conditioning, calibration and data quality
• MIL / SIL / HIL testing
• Test specification, verification and acceptance testing support
• Test benches, field and vehicle testing support