Extended Reality for Vehicles - Industry analysis

Extended Reality for Vehicles - TimeLine.

timeline
    title Extended Reality Development for Vehicles
    section 2021
      Oct : Meta & BMW partnership
    section 2022
      Nov : HoloRide & Audi
      Dec : DeepMirror Car XR
    section 2023
      May : Meta & BMW Quest Working In A Moving Car
      Oct : BMW metaverse
    section 2024
      Jan : Apple Vision Pro (Travel Mode)
      May : Meta Travel Mode
      Aug : DeepMirror Car XR update

Detailed TimeLine:

2021 Oct : Meta & BMW announced a partnership on XR-in-car problem. ¹
- Problem of Non-inertial reference frame: cameras and Imu are in conflict.
2022 Nov : HoloRide €700 kit, contains HTC’s ultra-compact Vive Flow headset and a wireless gamepad (€50) ². youtube video.
- Technology (3DOF + Canbus + recenter):
  - using only IMU for giving 3DOF movement (rotation only).
  - need user interface to recenter the screen (to driving direction).
  - (Audi only) uses the car’s real steering, accelerating, and braking as virtual movement.
2022 Dec : DeepMirror Car XR ³.
- Enable localization in moving vehicles, combined with car’s precise 6DOF.
- We could output : (1) 3dof w.r.t. car; (2) 6dof w.r.t world.

2023 May : Meta & BMW Quest Working In A Moving Car ⁴ ⁵.
- collaborated with BMW to incorporate IMU data from a BMW car’s sensor array in real time into the tracking system of XR devices.
- Next step : add the car’s location relative to the world, using the car’s precise 6DOF positioning system.
- not yet ready for customers.

2023 Oct : BMW metaverse ⁶.
2024 Jan : Apple Vision Pro ⁷ Travel Mode.
- Cannot work properly on moving vehicles.
2024 May : Meta Travel Mode with 6DOF⁸. It enables Quest to run in a moving car (where AVP’s travel mode failed to work).
2024 Aug : DeepMirror Car XR incar-6dof offering 6dof w.r.t. both car and world.

Some analysis

Meta’s develop path : switch from imu based slam system to pure vision slam system, to overcome the “Non-inertial reference frame” problem. This might be a good method, reducing the complexity of the system without tightly coupling with car’s imu.

Deepmirror’s develop path : we had moved faster than META, but with limited resources, have fewer hardware support. It took us longer to reach the pure vision method (same end).

Current trends: few customers are willing to try VR in moving vehicles, this market might has limited flexibility.

Key Assumptions of “Extended Reality for Vehicles” :

Technology is achievable. ✅
Customers are willing to use VR in vehicles. 🤔

References:

Share your opinion with me in Meta Community Forums, XiaoPeng Forums.

Others :