SpearEye: The direct solution for automotive safety through Location Awareness
1. Localization using GNSS/HD-GNSS/future 5G Rel 16 ground antennas
2. Location Awareness using dynamic HD-map display through localization and LTE-V2X location sharing capabilites.
3. Safety: speed limitation testing, Collision detection using HD-map lane level path planning.
Automotive visual sensors have finished their era, AI which uses neural networks recognizes only what it was trained for without "intuition", cameras works badly in dark light/rain/direct sun, Lidar works badly in rain, and uses AI. The only sensor that works in any weather and doesn't use AI is radar that will be used for collision detection.
A new way of thinking was needed to gain the Situational Awareness around the vehicle that works in any weather and recognizes objects around the vehicle even if there is no Line Of Sight to the object. Welcome to the Smart City era.
In Smart Cities we talk about using localization which is calculated using longitude/latitude on earth, but instead of calculating it using GPS/GNSS that uses satellites and it's accuracy depends on weather , we use HD-GNSS which fixes GNSS localization with weather data and soon 5G NR release 16 which uses ground antennas, does not depend on weather, and can reach accuracy of ~10Centimeters(2-3 Inches), adding to it the broadcasting/receiving capabilities of the 5G C-V2X modem, which broadcasts the position of the host vehicle which already has its own localization to the environment in ~1Km radius(0.5 mile) we get what we see in the simulation in the above bellow, which can be defined as next "Google maps"/"Waze"/"HERE WeGO" routing application with HD-map lane coordinates and multi object on same HD-map for Situational Awareness around vehicle, which is the next evolution of routing applications.
The simulation in the film above, runs real code that uses real HD-map lane coordinates in Berlin/Germany, with several vehicles, each vehicle calculates its own localization and broadcasts it to all vehicles around using the 5G C-V2X modem capabilities, all code is run within the modem, the modem can easily be the next "ECU" or the the computing engine that is in charge of safety and moves the vehicle.
To conclude the 3 steps into automotive smart cities:
Only after we have localization and created the Situational Awareness around vehicle, then we reach the safety level needed for collision detection, and only then we can talk about autonomous movement and platooning.
SpearEye develops the software that integrates all of the above.
We have been accepted into Plug and Play accelerator batch 9, Winter program, for Smart Cities/ Mobility in the Silicon Valley. We are looking forward to talking & meeting with PNP partners
Location Awareness: getting the full picture in real time around vehicle using best localization GNSS/HD-GNSS/5G NR that does not depend on weather, best HD-maps lane coordinates and best C-V2X broadcasting/receiving wireless communication capabilities direct between vehicles.
Safety based upon HD mapping & V2X modem localization and communication for collision detection.
Visual sensors integration as secondary system for recognizing non V2X objects around vehicle.
The vehicle interacts with environment & other vehicles for safety & services.
We confront the long transition and co existence period.
In this short video, we see two cars navigating toward destination and turn in junction, while choosing the correct lanes. We use Waze/ Google maps/ HERE WeGo application, that includes lanes. AV's will move using lane coordinates on HD-map.
We collaborate with HERE Technologies, to display dynamic HD-map of host car and other V2V enabled cars around (soon we'll present all vehicles around car, even without V2V modem, on HD-map).
Destination address is defined (like we do in Waze/ Google maps/ HERE WeGO), the AV receives a route to destination address, and use the HD-map lane coordinates to drive the car using exact lane coordinates from HDLM (High Definition Live Map). The AV defines its azimuth and speed according to its location inside road and uses speed limitation layer from the HD-map.
WE believe that HDLM will be the main enabler for driving and maintaining cruise, especially inside junctions and in bad weather, or without markings on road.
The dynamic HD-map bellow loads and parses a tile from Berlin,Germany that has ~900 roads/links and ~27,000 lane segments, and two vehicles taking different turns using its lane coordinates.
This simulatmonstrates dynamic HD-map display with HD-map lane coordinates showing roads, lane groups, lanes and junctions. Each vehicle uses smart cruise control to receives full global path plan from address to address,driving inside lane using HD-map lane coordinates in any road and in any junctions.
Localization is achieved by V2V modems that use latest capabilities.
Using HD-map, the vehicle can move in any lane including inside junctions, where no visual lane markings exist.