Reducing unplanned workshop visits
A European logistics group with more than 2,000 connected vehicles uses structured workflows to cut its roughly 180 unplanned workshop visits per day.
- Logistics
- IT-Service-Management
Key metric
Unplanned workshop visits
180115cases
Average unplanned workshop cases per day
At a glance
Starting point
The logistics group operates a Europe-wide fleet of more than 2,000 connected vehicles. Around 3,600 trips run every day, closely tied to telematics, on-board computers and the transport management system. At the same time, IT service management logs about 260 tickets a day on vehicle IT, telematics and workshop connections. On average, around 180 unplanned workshop visits occur each day, interrupting trips and tying up replacement vehicles. Information on causes and risks is scattered across telematics alarms, workshop emails, TMS notes and IT tickets.
- 2,000+
- connected vehicles
- 180
- unplanned workshop cases per day
The challenge
Unplanned workshop visits arise from a mix of technical faults, operating errors and delayed maintenance. Telematics systems generate hundreds of alarms every day, but only some of them are truly critical for the trip. Workshops report findings by email or phone, which are later transferred manually into tickets or TMS notes. Fleet control decides under time pressure whether a vehicle is taken off the road immediately or can keep driving. IT, the control center and the workshop rarely share the same information. Decisions depend heavily on individual people and are hardly ever documented systematically.
Typical symptoms in the organization
Why this remained unsolved
The problem had been known for years but was considered an unavoidable consequence of running a complex fleet. Earlier improvements focused on better dashboards in the telematics system and additional reports from the ticketing system. These approaches helped with transparency but did nothing to change the fact that decisions on workshop visits were made under time pressure and with scattered information. A classic IT project for full automation seemed too risky, because many individual cases require human judgment and rigid rules quickly reach their limits.
Our solution
Together with fleet IT, the control center and workshop coordination, an agentic workflow was introduced that bundles signals from telematics, the ticketing system and the TMS and delivers structured suggestions for handling potential workshop cases. Incoming alarms and fault reports are automatically assigned to a vehicle, a trip and an existing or new ticket. A digital assistant assesses urgency based on past cases, technical descriptions and current trip data and suggests actions, such as continuing with monitoring, a scheduled workshop visit at the end of the day or an immediate stop. Employees can adjust and justify suggestions and trigger follow-up actions in just a few clicks.
Where generative AI comes in
Generative AI analyzes free text from telematics alarms, workshop findings and tickets, summarizes the relevant information on a case and writes action suggestions in plain language. It suggests priorities and actions based on similar cases and creates structured notes for the TMS and the ticketing system.
- Connectors to the telematics portal, ticketing system and transport management system that bring vehicle, trip and fault reports together
- An agentic workflow continuously assesses new signals, bundles cases, suggests priorities and prepares follow-up actions
- An assistant interface for the control center and fleet IT that provides a consolidated view of each vehicle case, with the reasoning behind the suggestions
From alarm to structured workshop case
Bundle signals
Telematics alarms, workshop emails and IT tickets are read in continuously, matched to vehicles and trips and merged into cases.
Assess risk
The workflow assesses cases based on history, trip context and technical descriptions, generates a summary and suggests priorities.
Steer actions
The control center and fleet IT review suggestions, adjust them where needed and directly trigger follow-up actions such as workshop appointments or trip changes.
Learn from cases
Closed cases are analyzed so that the assistant gradually improves its future suggestions and risk assessments.
Measured: after six months in production
Period: after six months in production.
| Metric | Value | Basis |
|---|---|---|
| Unplanned workshop visitsKey metric | Down by around 36 percent (from 180 to 115 cases) | Time tracking before/aftermeasured |
| Processing time per case | Reduced from 40 to 25 minutes | Time tracking before/aftermeasured |
| Assistant suggestions adjustedQuality metric | Around 20 percent of suggestions are adjusted | Sampleestimated |
Cost-effectiveness
The solution significantly reduces unplanned workshop visits but does not replace technical diagnosis. Misconfigured telematics or missing sensor data limit its effect and require a deliberate sign-off by the control center or fleet IT.
Integration and operation
Transferability
The agentic workflow shows how signals from different systems can be bundled into a structured decision-making process without replacing human responsibility. A similar approach can be used wherever technical alarms, free-text reports and operational decisions come together. It is important that generative components explain their suggestions transparently and flag uncertainties, so that specialists can review them in a targeted way instead of relying on rigid rules.
A similar case in your organization? Let's talk.
Briefly describe your process. We'll get back to you with an initial assessment of whether the pattern can be transferred and what it would realistically take.