Controllab has over 25 years of experience in the design, implementation and testing of control software for complex and advanced systems. The heart of these control systems are generally closed loop control systems. We use simulation models and all the classical (time domain, frequency domain, PID, Kalman filtering,…) and modern (adaptive control, machine learning,..) control techniques to make these control systems stable and efficient.
The key to high quality control software is test automation. We have developed a toolkit for the automated testing control software:
- Testing stability and efficiency with simulation models.
- Unit testing and functional testing in the software development system.
- Functional testing on the controller hardware (PLC, PC, embedded), HIL simulation.
With machine learning you can automate the work that an expert would do:
- Run a machine or system and record data
- Process the data
- Detect causes and effects
Controllab has the expertise to implement machine learning based on time series data. We use existing data sets to train a network to inspect the data as an expert would do. If the data is not available, we can use simulation models to generate data sets.
Controllab has build up expertise for several applications.
- Cranes: motion compensation, anti-sway, tag line control, …
- Access Bridges: motion compensation, automated landing, collision detection, ..
- Monopile Grippers: Tilt control, motion control, overload protection, …
- Dredging: motion compensation
- Robotics: motion control