Sustainable & digital engineering Advanced materials

LOCAFLEET: real-time state estimation algorithm with a 100% guarantee

Fill out the form to download the document

Established state estimators (Extended Kalman Filter, Unscented Kalman Filter) calculate probabilistic position, velocity or acceleration zones. The approximate nature of the assumptions (error distribution of prediction models and sensors) on which these methods are based may invalidate the probability zones calculated by these methods. This raises safety issues for critical applications such as collision avoidance and fleet management

COMPETITIVE ADVANTAGES

  • 100% guaranteed
  • Not highly conservative
  • Real-time
  • Easy to integrate

APPLICATIONS

  • Space: Satellite constellation control, debris avoidance, etc.
  • Drones: Fleet management for training flights, Geofencing, etc.
  • Aeronautics: Air Traffic Management, Collision Avoidance System, etc.
  • Automotive: Automatic braking and collision avoidance system, etc.

INTELLECTUAL PROPERTY

  • Patent

DEVELOPMENT STAGE

Laboratory validation of the technology

LABORATORIES

Logo laboratoire LAAS CNRS
  • LAAS CNRS

Description

The technological solution that has been developed is based on set-based estimation to provide location zones and guaranteed speed or acceleration ranges in real time.

The key points of the current system are:

  • Algorithm providing a 100% guarantee that the actual state of the observed system is within the estimated range
  • Not highly conservative, “real-time” estimation
  • Simple integration into a system architecture, as it can be implemented on a conventional embedded board and uses existing sensors
  • Suitable for non-linear systems
  • Suitable for observation delays and uncertainties in observation times

 

image présentant l'offre locafleet

Technical specifications

  • Probability of results: 100% guarantee
  • Implementation: Can be embedded on conventional boards
  • Processing frequency: 1 Hz (in its current version)