Location Awareness: Understanding the surrounding around the vehicle using HD-map with localization&broadcast capabilities of V2X modem
Barak Rosenberg
I am an accomplished AI specialist, founder, entrepreneur, manager, programmer, architect. With a rich background in the hi tech industries, management, architecture, programming, AI/ML (including LLM edge/cloud).
In automotive location awareness, we mean the vehicles builds all the dynamic objects that include vehicles and pedestrians around the vehicle using HD-map lanes, calculates it's own location using GNSS/5G NR high band ground antennas to the level of 6th digit of longitude/latitude by calculating Time Of Arrival of packets from ground antennas, broadcasting host location: longitude,latitude,speed,heading to all vehicles around using C-V2X protocol, receiving location from enabled V2X vehicles and pedestrians and be able to localize them on the HD-map which includes lane coordinates.
This is what we get:
In the film above, in the white lines bellow we can see location data with precision of 6th digit ~10cm/~4Inch sent from V2V modems to HD-map dynamic display, heading with precision of 3rd digit, lane data from HD-map, road data from HD-map, speed limitation taken from HD-map speed attributes layer compared to actual speed of vehicle, movement based on global route plan integrated with local path plan inside lane for real time movement in <50ms. The vehicle doesn't need to "see" or recognize each other, data is shared between vehicles, sending data and receiving data, peer to peer takes 2-3ms, and the technology works in any weather and without Line Of Sight between vehicles, each vehicle tests its own speed compared to HD-map speed limitation data and its location, as well as other vehicles around the vehicle at a radius of ~1Km which is C-V2X protocol broadcasting capabilities, we can say that each system acts as a police system that checks its own speed and vehicles around in real time. Again algorithms using V2X localization and broadcasting capabilities with HD-map lane data work in any weather, no need for Line Of Sight, using MEC(Mobile Edge Computing) in real time <50ms, no need to do upload/download to send data to cloud and back, each vehicle can recognize vehicle speeding or moving strange in ~1Km/~0.6Mile radius which can be even 3 streets around the vehicle without the need to "see" the vehicles or pedestrians moving there.
领英推荐
Another example:
In this example we can see collision detection algorithm between vehicles using path planning inside lane over HD-map, each vehicle tests its own possible collision with other objects which can be vehicles/pedestrians.each vehicle can test other vehicles possible collision with other vehicles/pedestrians. We are showing capabilities like recognizing a human moving into the road by several vehicles and pedestrians, recognizing it from vehicles several streets away up to 1Km away from the pedestrian, the pedestrian or vehicle can know about possible collision from location awareness software in its own phone, or can receive multiple warning from other vehicles/pedestrians around.
We are talking about swarm technology in real time testing and calculating movement in <50ms, not only of our own host vehicle, but other vehicles and pedestrians around, calculating movement and testing movement in real time of multiple objects, using parallel computing algorithms.
I Help Tech companies transform their vision into paying products. Proven success with $100M+ Industry Leaders, Align your product with customers and investors in 90 days
2 个月???? ??? ?? ?? ???????? ??? ????? ???? ?????? ???: ?????? ????? ??? ??????? ????? ????? ?????? ??????. https://chat.whatsapp.com/BubG8iFDe2bHHWkNYiboeU