open scenario map

open scenario map

open scenario map

open scenario map

Open scenario
map

Automotive safety validation — open sourced

unseen scenarios at scale

unseen scenarios at scale

Scaling the test of generalization for ADAS systems beyond the development fleet.

2021
2021

YEAR

1
1

Cities

1
1

Experts

893
893

Incidents

342
342

Severe incidents

4839
4839
4839

KM

scenarios that matter

We partnered with driving school instructors to annotate scenarios where human drivers make unsafe driving decisions, or critical errors. The labels are categorized by the type of driving mistake; e.g. 'Right of Way' for not following the appropriate priority order at an intersection, or 'Placement' for incorrect placement of the vehicle between lane markings. Driving intructors additionally annotate severe driving mistakes by using dual-control pedals. These scenarios are labeled as 'Instructor Pedal'. See the current cohort of scenarios annotated by driving instructors in the interactive map on the right.

scenarios that matter

We partnered with driving school instructors to annotate scenarios where human drivers make unsafe driving decision or critical errors. The labels are categorized by type of driving mistake. F.ex Right of Way for not following the appropriate priority order at intersection or Placement for incorrect placement of vehicle between lane markings. Driving intructor additionally annotate severe driving mistakes by using dual control pedals. These scenario are labeled as Instructor Pedal.

scenarios that matter

We partnered with driving school instructors to annotate scenarios where human drivers make unsafe driving decisions, or critical errors. The labels are categorized by the type of driving mistake; e.g. 'Right of Way' for not following the appropriate priority order at an intersection, or 'Placement' for incorrect placement of the vehicle between lane markings. Driving intructors additionally annotate severe driving mistakes by using dual-control pedals. These scenarios are labeled as 'Instructor Pedal'. See the current cohort of scenarios annotated by driving instructors in the interactive map on the right.

scenarios that matter

We partnered with driving school instructors to annotate scenarios where human drivers make unsafe driving decisions, or critical errors. The labels are categorized by the type of driving mistake; e.g. 'Right of Way' for not following the appropriate priority order at an intersection, or 'Placement' for incorrect placement of the vehicle between lane markings. Driving intructors additionally annotate severe driving mistakes by using dual-control pedals. These scenarios are labeled as 'Instructor Pedal'. See the current cohort of scenarios annotated by driving instructors in the interactive map on the right.

scenarios that matter

We partnered with driving school instructors to annotate scenarios where human drivers make unsafe driving decisions, or critical errors. The labels are categorized by the type of driving mistake; e.g. 'Right of Way' for not following the appropriate priority order at an intersection, or 'Placement' for incorrect placement of the vehicle between lane markings. Driving intructors additionally annotate severe driving mistakes by using dual-control pedals. These scenarios are labeled as 'Instructor Pedal'. See the current cohort of scenarios annotated by driving instructors in the interactive map on the right.

scenarios that matter

We partnered with driving school instructors to annotate scenarios where human drivers make unsafe driving decisions, or critical errors. The labels are categorized by the type of driving mistake; e.g. 'Right of Way' for not following the appropriate priority order at an intersection, or 'Placement' for incorrect placement of the vehicle between lane markings. Driving intructors additionally annotate severe driving mistakes by using dual-control pedals. These scenarios are labeled as 'Instructor Pedal'. See the current cohort of scenarios annotated by driving instructors in the interactive map on the right.

Vektorized

'Vektorized' scenarios are first captured with eight HDR cameras, and then morphed by Vektor into the OpenScenario format for an unbiased, unseen test of generalization for ADAS systems in simulation.

Vektorized

'Vektorized' scenarios are first captured with eight HDR cameras, and then morphed by Vektor into the OpenScenario format for an unbiased, unseen test of generalization for ADAS systems in simulation.

Vektorized

'Vektorized' scenarios are first captured with eight HDR cameras, and then morphed by Vektor into the OpenScenario format for an unbiased, unseen test of generalization for ADAS systems in simulation.

Vektorized

'Vektorized' scenarios are first captured with eight HDR cameras, and then morphed by Vektor into the OpenScenario format for an unbiased, unseen test of generalization for ADAS systems in simulation.

Vektorized

'Vektorized' scenarios are first captured with eight HDR cameras, and then morphed by Vektor into the OpenScenario format for an unbiased, unseen generalization test for ADAS systems in simulation.

Advancing autonomy through test for generalization

fleet & hardware
  • Expand coverage by sponsoring a driving school

  • Bridge data gaps by completing drive missions

Computing
  • Donate cloud credits for hosting Open Scenario Map

  • Donate GPUs for scenario extraction with Vektor

data & MAPS
  • Contribute lane-level maps to expand Open Scenario Map

  • Add natural language commentary to scenarios

research & Development
  • Partner with Yaak to advance the development of Vektor

  • Contribute to Open Scenario Map by reviewing scenarios

Join the waitlist

Get notified when Open Scenario Map is public.

Join the waitlist

Get notified when Open Scenario Map is public.

join the waitlist

Get notified when Open Scenario Map is public

Join the waitlist

Get notified when Open Scenario Map is public.

Join the waitlist

Get notified when Open Scenario Map is public.