beta version

Beyond Telemetry

Data Fusion for new era Telemetry

Hetereogeneous systems such as the one you could find in F1 car, generate heterogeneous dataset.

Mechanical stresses and extreme conditions during the race could corrupt these data. Futhermore some kinds of signals such as audio and video require specific data processing pipelines. Our state of art computer vision algoritmhs are able to mix this source creating predicitve models.

The platform designed supports race directors with a all in one application: on this monitor they can manage any complex situation.
  1. On-board Camera Tagging
  2. Use Deep Learning based classifier to recognized interesting events using only data from rear/front cameras
  3. External Camera Tagging
  4. Use Deep Learning based classifier to recognized interesting events using only data from external cameras
  5. Telemetry based frame tagging
  6. Exploit telemetry data to label frames with a more accurate model.
  7. Real Time Frame based
  8. Ranking tagged frame to expose most informative ones using only camera data
  9. Predictive Frame based
  10. Ranking tagged frame to expose most informative-to-be ones using only camera data
  11. Real Time Telemetry based
  12. Enrich the real time ranking algorithm using features extracted from telemetry data
  13. Predictive anomaly detector module
  14. Enrich the prediction ranking algorithm using features extracted from telemetry data Enrich the real time ranking algorithm using features extracted from telemetry data
  15. 3D car models
  16. Create a 3D model of the car from camera cars to create artifical panoramic view of the car from different point of views (drones-like camera)
  17. Trajectories projections
  18. Project actual car trajectory over the video as an “arrow” of the direction the car is going to follow
  19. Real time social fresh trends
  20. Write a ranked shortlist of trendy tweets/fb posts/topics
  21. RT widgets
  22. Create more visual widgets exploiting new information (e.g., represent wheel speed over a 3d representation of the wheel itself)