dRISK Announces its Launch

dRISK aims to make autonomous vehicles as safe as possible as quickly as possible.  The new patented knowledge graph technology (analogous to Google's knowledge graph of the internet but in dRISK's case a knowledge graph of real-world events) solves several problems which have plagued AV developers so far.

FREMONT, CA: dRISK, a London and Pasadena-based company that has been in stealth mode until now, has announced its launch and that it has used its edge case retraining tool to achieve a 6x performance in time to detect high-risk events for Autonomous Vehicles (AVs) for the first time commercially.

Currentlysemi-autonomous and autonomous vehicles do not often detect high-risk events in time to react to them (oncoming cars peeking into the lane from behind other vehicles, vehicles running red lights hidden by other vehicles). dRISK's tools for retraining autonomous vehicles to recognize edge cases represent a significant step forward in the ability to retrain autonomous vehicles to outperform humans. On April 12, 2021, the findings were formally presented at NVIDIA's GTC meeting.

dRISK aims to make autonomous vehicles as safe as possible as quickly as possible.  The new patented knowledge graph technology (analogous to Google's knowledge graph of the internet but in dRISK's case a knowledge graph of real-world events) solves several problems which have plagued AV developers so far. It helps encode enormously high-dimensional data from all the different relevant data sources into a tractable form, and then providing the full spectrum of edge cases to retrain on not just with what has already occurred but will happen in the future.

To achieve superior testing and retraining outcomes for customers on real-life data, dRISK delivers simulated and real+simulated edge cases in semi-randomized, impossible-to-game training and test sequences. dRISK's edge cases are trained to identify only the predictors of high-risk incidents instead of conventional training and development methods, which train AVs to recognize whole vehicles and pedestrians under favorable lighting conditions (for example, the headlights of an oncoming car peeking into the lane amid low-visibility). AV systems that have been trained in this manner can detect high-risk events faster while maintaining a high level of performance on low-risk events.

AV developers, mass transit authorities, and one of the world's largest insurers are among dRISK's customers, all of whom are interested in reducing AV risk and improving AV efficiency. dRISK has so far only provided semi-customized solutions on an individual basis, but it plans to release a web-based version of its AV retraining product later this spring to make these capabilities available to the broader AV community.

The UK's Centre for Connected and Autonomous Vehicles awarded dRISK Inc and its wholly-owned subsidiary dRISK.ai Limited 3 million pounds to build the ultimate driver's test for self-driving vehicles, along with its partners in the UK-based D-RISK consortium, Imperial College London's Transport Systems Laboratory, Claytex, and DG Cities.