How AI and connected diagnostics can help fleets predict faults and speed up repairs

September 7, 2026

If reducing downtime is the objective, the challenge for fleets is increasingly how to turn growing volumes of vehicle data into faster, better maintenance decisions.

Benoit Dessart, Head of BL Digital Solutions at ZF Aftermarket explains how connected diagnostics, AI and predictive maintenance are changing the repair process – helping workshops identify faults earlier, prepare parts and repairs in advance and move towards a more proactive model of vehicle maintenance.

What software capabilities should fleet managers be looking to better understand in order to cope with the influx of this vehicle information?

Using edge computing, vehicles process and filter this information in real time, transmitting only the most relevant data to manufacturers and fleet management platforms. Diagnostic platforms can then interrogate thousands of parameters in seconds, enabling workshops to begin investigating faults before a vehicle even arrives.

AI is accelerating that transformation and that is raising the bar for workshops where efficiency is vital to their business. Various studies cite that AI-driven predictive maintenance can reduce downtime anywhere between 20-50% across vehicle environments by enabling maintenance to be scheduled before failures occur.

While traditional diagnostic tools identify faults after a failure, machine-learning systems can analyse voltage, vibration, temperature and other operating data to identify faults or deteriorating components triggering a Diagnostic Trouble Code (DTC) – an alphanumeric code generated by a vehicle’s onboard computer to identify a specific malfunction or irregularity in the vehicle’s systems.

Walk into a modern commercial vehicle workshop and the first tool a technician reaches for is often a diagnostic tablet rather than a spanner. This is a simple sign of how far the aftermarket has changed.

What digital difficulties should fleet managers be aware of when it comes to minimising downtime?

Modern diagnostic systems are exceptionally effective at identifying abnormal behaviour, but less effective at explaining why it has occurred. When a technician connects a diagnostic platform to a vehicle, the system retrieves DTCs from the ECUs. Retrieving a DTC typically takes only a matter of minutes. Determining why that code appeared, however, can require several hours of data analysis, electrical testing and physical inspection.

A wheel-speed sensor fault, for example, might ultimately be caused by damaged wiring, connector corrosion, mechanical wear, a software issue or the sensor itself. The fault code narrows the investigation. It does not complete it.

Diagnostic Trouble Codes are digital clues, not definitive answers. A code only indicates a sensor crossed a threshold, not whether the cause is a failed component, corroded wiring or something else.

Software highlights anomalies, but human expertise is required to synthesise data with physical context to find the true root cause.

How can connected vehicle data and remote diagnostics help fleets and workshops identify faults earlier, plan repairs more effectively and ultimately reduce vehicle downtime?

Connected vehicle technologies allow workshops to begin diagnosing faults remotely while vehicles remain in service. Parts can be identified before arrival, schedules adjusted automatically and customers kept informed throughout.

ZF's digital technologies support this connected workflow. Its ZF [pro]Diagnostics platform, incorporating ZF MultiScan, enables fleets to perform multi-brand diagnostics across passenger cars, light commercial vehicles, trucks, trailers and buses. ZF Bus Connect, CAN Tunnel, ZF SCALAR and ZF [pro]Manager extend those capabilities into remote vehicle monitoring, workshop management, fleet analytics and digital customer communication.

Together, these technologies allow fault detection, diagnosis, parts identification and repair planning to operate as one continuous digital process, rather than disconnected tasks. Recognition of our approach came in 2025, when ZF [pro]Manager received the Equip Auto Innovation Award for Digital Solutions and Connectivity, recognising its contribution to improving workshop efficiency and customer communication.

How do you expect AI and predictive diagnostics to change workshop operations over the next decade, and where will skilled technicians continue to add value that technology alone cannot replace?

Machine-learning algorithms will become increasingly capable of analysing millions of historical repair records, identifying emerging failure patterns and recommending the most probable repair pathways almost instantly, allowing technicians to focus on diagnosis, validation and repair. Predictive maintenance and remote diagnostics will also become more sophisticated, speeding up the entire process and reducing downtime – the ultimate aim.

As vehicles become increasingly connected, the volume of diagnostic information available to workshops will continue to grow. Artificial intelligence will become better at recognising patterns, predicting failures and recommending repairs, while connected ecosystems will make the repair process faster, more efficient and more proactive.

Yet the objective remains unchanged: keeping vehicles on the road. Digital diagnostics, predictive maintenance and AI are becoming indispensable tools, but they achieve their greatest value when combined with the judgement, experience and engineering expertise of skilled technicians.

Summary

For fleet operators, the value of connected diagnostics and AI will ultimately be measured in something very simple: whether they help keep vehicles productive and on the road.
The technology is becoming increasingly capable of identifying patterns, predicting failures and accelerating diagnosis, but Benoit’s view is that the strongest results will come from combining those capabilities with skilled technicians who can interpret the data, verify the diagnosis and complete the repair correctly.