On a dispersed charging network the dominant service cost is getting a person to the site, not the work done once there. That single fact should shape the entire service model.
It means the highest-value investments are the ones that avoid a visit, or that make the visit succeed the first time.
Diagnose before dispatching
A meaningful proportion of reported faults can be resolved remotely or need no action at all: vehicle-side refusals, transient communication problems, conditions a reset clears.
Attempting remote resolution before dispatch is the largest single reduction in service cost available, and it requires diagnostic depth rather than only fault notification.
First-time fix depends on what the engineer knows before arriving
An engineer dispatched with a code and an address will sometimes diagnose and return, which doubles the cost of that fault.
- The specific fault and any vendor-specific detail, more than a standard code.
- Firmware version and recent update history.
- Whether this unit has failed this way before.
- Site access details, which stop more visits than technical problems do.
- The part most likely required, so it travels with the engineer.
Site access is underestimated
Barriers, permits, escorts, restricted hours and keys held by someone unavailable cause more failed visits than technical difficulty. This information belongs in the site record and needs to be current.
Recording it after the first failed visit is the common pattern, and recording it at commissioning is far cheaper.
Spares strategy follows lead time, not failure rate
Parts with long lead times or single sources need deeper cover than their failure rate alone suggests, because the exposure is the wait rather than the frequency.
Holding a small stock of long-lead items locally frequently costs less than the downtime one shortage causes.
Batch what can be batched
Where travel dominates, combining a fault visit with preventive work, inspections and pending firmware attention on nearby units converts one journey into several outcomes.
This requires knowing what is pending per site, which is a data problem rather than a scheduling one.
Close the loop into engineering
Field failures are the highest-quality information a product programme receives and the hardest to collect. A service operation that closes tickets without recording root cause converts that information into nothing.
Requiring a cause at closure, from a defined list, is what makes the same fault stop recurring rather than being fixed repeatedly.