Bharat in Betastartup field guide
← Back to the atlas
No. 1396Problem statementMobility and logistics

AI-ML based Intelligent Dead Reckoning system for seamless navigation

Source

Source summary

Develop a smartphone app and edge engine maintaining vehicle navigation through GNSS outages using AI/ML IMU dead reckoning and fusion, without OBD/speedometer wiring. Calibrate phone alignment, filter vibration/potholes, infer velocity, constrain trajectories with offline roads and transition back to GNSS seamlessly. Train before deployment, demonstrate IO-VNBD-based preliminary results and support external IMUs as well as phones. Targets include drift below 10% of distance during blackout, 10 Hz phone updates and around 200 Hz with suitable FOG-IMU edge input. Lane-level language is an aspiration, not established accuracy.

No. 417Problem statementMobility and logistics

Indigenous development of door control unit (DCU) and gear motor for automatic operation of platform screen doors at metro stations. these platform screen doors are for passenger safety and isolates platform from track.

Develop platform-screen-door control units and geared motors for metro stations, with SIL3 safety requirements. The system must detect obstacles, adjust door timing and synchronise left/right doors to maintain the barrier between passengers and tracks.

No. 427Problem statementMobility and logistics

Design of a steering console for an underwater vessel

Develop an underwater steering console integrating vehicle-motion measurements and control information. The source includes depth, speed, heading and trim, real-time serial inputs and manual, automatic and emergency modes. Detailed control implementation is omitted from this military-system overview.

No. 428Problem statementMobility and logistics

Design of electro-magnetic velocity log for navigation of underwater vessels.

Develop an electromagnetic velocity log for underwater navigation, supporting operating and simulation modes. The visible source gives a basic equipment requirement rather than a detailed user workflow or demonstrated measurement performance.