To enhance air transport safety, the main objective of VISION is to validate smarter technologies for aircraft Guidance, Navigation and Control (GN&C) by including:

  1.  vision-based systems,
  2.  advanced detection and resilient methods.

With respect to the first point, and as the name suggests, one of the key features of this VISION project is the utilization of on-board vision systems. From the perspective of detecting and handling flight anomalies, usually the flight control performance failure is detected by Fault Detection and Diagnosis (FDD) techniques using proprioceptive sensor measurements. However, if the problem comes from an aircraft physical system failure such as actuator jamming, this can also be detected by visually monitoring the aircraft. Based on this idea, the VISION project proposes a vision-based control surface monitoring system which provides additional information or confirmation, and will be fused with the more classical FDD techniques. In addition, VISION will investigate vision-aided local precision navigation systems which covers poor GNSS integrity. In the case of accuracy degradation in the GNSS signals, the vision-based measurements can be fused with the degraded GNSS information to maintain precision in the navigation performance. At the same time, visual sensing technologies and image processing algorithms will be investigated, that can provide the required vision information accurately and robustly even in degraded visibility conditions.

With respect to the FDD/FTC techniques studied in the second point, VISION will build on the results from the precursor FP7 projects (ADDSAFE and RECONFIGURE) as well as the equivalent Japanese projects (AFCG and IFTC). These techniques include H-infinity, LPV, sliding mode, on-line parameter tracking, and adaptive control. But in opposition to the previous projects, VISION focuses on their consolidation for real-time implementation, practical feasibility, and on the effective matching of the different FDD and FTC algorithms developed by all partners.