TEKIA: supporting bus maintenance with AI

A technical assistant for maintenance teams

TEKIA helps technicians find information and diagnose faults. The assistant draws on manufacturers’ documents, Keolis procedures and teams’ field experience. It makes these resources easier to access through a shared knowledge base.

This approach addresses practical needs:

• finding a procedure,

• understanding a fault,

• consulting guidance relevant to a maintenance task.

TEKIA aims to reduce diagnostic time and prevent repeat repairs for the same fault. This approach is part of a broader effort to improve vehicle reliability and support predictive maintenance, helping the network run reliably. For transport authorities and passengers, it helps make buses more reliable and ensure continuity of service.

Three support options tailored to workshop work

Technicians can interact with TEKIA using text, images or voice. These three options suit the different situations they encounter on the job:

1. Text-based interaction makes it easier to search technical documentation.

2. Image analysis helps identify a part, a risk or an issue from a photo.

3. Voice guidance lets technicians access advice while keeping their hands free and their attention on the task.

During an inspection, image analysis can help interpret a warning light, check expiry dates or spot unusual wear. The assistant supports the technician, who remains central to the diagnosis and repair.

Shared knowledge to pass on expertise more effectively

TEKIA makes it easier for maintenance teams to share their experience. Insights from the field enrich the shared knowledge base and help teams identify recurring faults. Information gathered during one repair can therefore help other technicians.

The innovation also supports the training of new employees. Guided, interactive learning paths help them carry out first-level diagnostics. The aim is to build their independence gradually and safely, while reducing the time experienced technicians need to spend supervising them.

A gradual rollout shaped by real-world use

The rollout is starting with two pilot depots, before a wider deployment is considered.

Early feedback points to several promising uses:

• quickly finding information in manufacturers’ documentation;

• analysing images to spot signs of wear or leaks, locate technical plates in the engine compartment and help troubleshoot faults;

• guiding technicians through complex fault resolution by setting out the steps and protocols to follow;

• inspecting vehicles before their roadworthiness test to identify issues that need attention in advance and help avoid a retest.

With TEKIA, Keolis is putting artificial intelligence to work in support of frontline expertise, helping improve the performance and safety of bus maintenance.