Operating design domains: defining where an automated service can work

An ODD is a boundary, not a slogan

UK Government guidance describes an automated driving system as operating within specific driving situations, sometimes called an operational design domain, or ODD. The conditions may be narrow or broad. Outside the stated situations, the system may need a different control arrangement or a safe response.

For a local service, the ODD should make the intended geography, roads, speeds, time of operation, weather, visibility, traffic context, connectivity assumptions and vehicle capability clear. The point is to define where the system has been designed and evidenced to work—not to imply that it can operate everywhere.

Sources: [1]

The boundary must be connected to evidence

The UK Government’s safety-principles work records that pre-deployment information should include the expected ODD and evidence that the system can operate safely in those conditions. It also identifies the importance of a safety case, testing results, adverse environments, vulnerable road users and lifecycle risks.

That means a local proposal should connect each claimed operating condition to a test, a control, a monitoring method or a clearly stated limitation. A map alone is not an operating case.

Sources: [2]

Local geography changes the design

A future local feasibility process should examine the actual roads, stopping points, access routes and customer handover locations. Narrow lanes, steep sections, poor surfaces, blocked routes, stairs, private paths and difficult building access may require a different operating arrangement or an exclusion from the defined service area.

This is a design inference for future feasibility, not a claim about an approved SaveMe service. It is also why an autonomous road vehicle should not be presented as a solution that simply travels everywhere a person might walk.

What this means for SaveMe Scotland

SaveMe Scotland can use the ODD concept to structure a partner-led local evidence process: define the user need, map the intended service area, identify safe pickup or handover points, specify exclusions, record incidents and test the operating response when conditions change.

SaveRide remains a proposed capability, not a live autonomous passenger service. No council has approved a SaveMe Scotland pilot or deployment at this stage, and any future service would still require suitable partners, safety evidence, funding, insurance and the applicable regulatory and local decisions.

Sources: [1] [2] [3]

References

1. [1] UK Department for Transport: Automated vehicle trialling code of practice
2. [2] UK Department for Transport: Automated vehicles: statement of safety principles — summary of responses and next steps
3. [3] UK legislation: Automated Vehicles Act 2024