The dRISK Knowledge Graph is a scenario database and risk assessment solution for AV developers, fleet operators and infrastructure planners.

Ensure test coverage over the entire space of risk

Analyze scenarios using comprehensive risk metrics

Use Cases

Explore how dRISK can help you build safer AVs

Ensure testing coverage over the space of risk

Gain immediate insight into the space of risk covered by your current tests, then plug any gaps by selecting the scenarios you need directly from dRISKs Knowledge Graph using our comprehensive taxonomy


Full geospatial projection and ODD coverage and Comprehensive risk metrics

Explore real or simulated scenarios which are statistically significant for a chosen Operational Design Domain, and view scenarios projected in situ


Scenario adaptation, creation and domain randomisation

Copy and modify scenarios, create programmatically or even manually; generate even more adversarial behavior, e.g. with dRISK’s custom recombination algorithm


Feature Space Exploration and failure mode identification

Multi-scale: assess scenario-by-scenario, drill down to specific behaviours, or zoom out to see the forest for the trees. Find sources of failure modes in minutes in simulation or real life, guaranteed faster than O(log(Ndimensions))


Comprehensive risk metrics

Dozens of industry-standard metrics (e.g. RSS) and thousands of additional metrics (e.g. at-fault, percent occlusion, total energy), fully programmatically available


Import, explore, and mine your own data

Be it scenario data, real-life recorded data, or a list of ODD feature observations, dRISK Edge enables customers to import and extract insights from vast amounts of heterogeneous data, all in one place


Automatic feedback for nightly regression tests

Automatically fuse regression tests on a periodic or ad-hoc basis into existing data, for weekly/daily/hourly monitoring performance. Periodically update embedding of all scenarios to reveal progress in new tests, export reports flexibly