EV Magazine April 2023 | Page 66

The research team at NVIDIA developed its Neural Reconstruction Engine , which takes imagery and video recordings from the road to feed data into the AI network , storing millions of driving scenarios .

AI automatically extracts the key data points required to activate simulations of the same scenarios , but are fully reactive and are manipulated as needed . When it comes to assembling the environment , AI converts 2D video data from a real-world drive into a 3D digital twin that can be used to inform the DRIVE Sim .

Experiencing autonomous driving firsthand has already helped some individuals understand the behaviour of the vehicle and , generally , feel at ease very quickly . Knowledge is also a key factor in understanding how safe autonomous vehicles are .

When put into perspective , the neural network that has been developed for autonomous cars more capable of correctly assessing risks on the road — not to mention the next generation vehicle-to-vehicle ( V2V ) connectivity on the horizon .

With cars talking to each other while also analysing tens of millions of data points in a 360-degree-view of the vehicle , potential disturbances or incidents are reduced greatly in comparison to the 20-20 vision of the healthiest human drivers .

Such solutions are being adopted across the EV landscape . One of the most recent applications of autonomy is with the Volvo EX90 — the latest and safest automobile to be built by the Swedish OEM .

With so many organisations shifting to electric or entering the industry on a pure-play basis , the vast playing field of technology capabilities from EVs is enabling more and more organisations to ready their cars for autonomous capabilities .

66 [ April 2032 ] EVMAGAZINE . COM