Organised Editing/Activities/aiMotive

From OpenStreetMap Wiki
Jump to navigation Jump to search

aiMotive

aiMotive (formerly AdasWorks) is an autonomous vehicle technology company. It works with automotive manufacturers and Tier‑1 suppliers to develop and deploy automated driving systems. The company describes its approach as “vision‑first”: the system relies mainly on cameras and artificial intelligence to perceive the environment. The technology is intended for integration by carmakers to build autonomous vehicles that can operate in a wide range of conditions and locations. In September 2017, the PSA Group (now part of Stellantis) entered a partnership with aiMotive. On 22 December 2022, aiMotive was acquired by Stellantis, the multinational automotive manufacturing corporation.

Website: aimotive.com

Timeframe

2026 February and March

Goals

Our goal is to improve OSM-based GPS and navigation by uploading correct, source-backed data. We focus on attributes such as speed limits, lane counts, and turn lanes so that the OSM basemap is more accurate and reliable for positioning and routing in the regions we edit.

Tools and data sources

We improve the map using data collected by our company vehicles. Our primary source is imagery from cameras mounted on these vehicles (street-level imagery). This footage is used to verify and update map features such as roads, traffic signs, crossings, and other relevant details. Edits are made in line with OSM practice: we only add or change data when it is clearly supported by this imagery, and we reference the source in our changesets (e.g. in the changeset comment and, where appropriate, in tags such as source=aimotive) so that our contributions are traceable and verifiable by the community.

In-house toolIn-house tool

We use an in-house developed tool to process the imagery from our vehicles and to prepare map edits. The tool supports reviewing multi-camera footage, comparing it with existing OSM data, and exporting or uploading edits in line with OSM conventions.

Availability of sources

Our data sources (vehicle-mounted camera imagery and any related internal data) are not publicly available. We always indicate the source in our changesets (e.g. in the changeset comment #aimotive ) so that our contributions are traceable and verifiable by the community.

Participants

Rationale

Correcting speed limit, lane count and turn:lanes metadata in US.

Results

During the planned two-month project period, the aiMotive team edited nearly 18,000 road segments in the project area. All edits were manually reviewed before upload and were made only when clearly supported by our street-level imagery.

Mapping work included geometry improvements, corrections to lane counts, updates to turn restrictions and turn lane tagging, and fixes to incorrectly tagged speed limits, especially where values had been entered using km/h instead of the locally appropriate mph units.

The project work was carried out in the following areas:

  • Delaware
  • District of Columbia
  • Maryland
  • Michigan
  • New Jersey
  • New York
  • Ohio
  • Pennsylvania
  • Virginia

This activity is limited to road-related mapping. We do not edit unrelated features such as addresses, POIs, buildings, or landuse as part of this project.

Contact

Organiser: aiMotive

Contact mail: levente.szeman@aimotive.com

Hashtag

We use a consistent changeset comment format so our edits can be easily identified and reviewed by the community. Where appropriate, the source is also referenced in the changeset metadata.

#aimotive