AGV fleet operation: KPIs, servicing and continuous improvement
Launching the fleet is not the end of the project but the start of its longest phase — day-to-day operation. It is in AGV operation that fleet KPIs, servicing discipline and work on continuous improvement determine whether the investment pays back within the planned time. This guide closes the "AGV implementation step by step" series: after the system has been handed over, it shows how to keep it running, what to actually measure and how to develop automation instead of leaving it to its own devices.
Post-implementation support and SLA — don't face it alone
The first days after start-up are when minor configuration problems come to light and the crew has the most practical questions. For a plant with no automation experience, the key thing is not to deal with this single-handed. Many suppliers offer intensive support during this period — a few days or weeks of a service engineer on site, or round-the-clock telephone availability. It is worth taking advantage of this: a specialist's rapid response limits downtime and gives employees the reassurance of having an expert on hand.
Before the start-up support ends, agree a standing working arrangement with the supplier: a channel for reporting problems (for example, a dedicated helpdesk) and escalation rules. If you have signed a service level agreement (SLA), treat it as a living document rather than something filed away in a drawer — monitor whether the service team arrives within the agreed response time and whether it solves problems effectively rather than just temporarily. This is data that will come in handy in every subsequent discussion about expanding the system.
Maintenance and servicing: inspections, batteries, wheels
Fleet reliability does not come from the quality of the equipment on the day of handover, but from the regularity of inspections. Plan and carry out periodic technical inspections — depending on the supplier's recommendations, usually quarterly and performed by the service team. They should cover the technical condition of the AGVs and AMRs themselves: sensors, navigation systems, battery wear and the wheels and castors, which are among the first components to wear out under intensive use. An inspection also covers the infrastructure around the vehicles — the cleanliness of navigation markers, the condition of the communication network — as well as any software updates.
Two practices significantly shorten the response time to a fault. The first is keeping a stock of basic spare parts — sensors, wheels, fuses — so that a replacement does not have to wait for a delivery. The second is training your own maintenance team well enough to carry out minor servicing tasks on their own. If the technician on shift replaces a worn part without calling in an external service provider, the vehicle is back on its route in minutes, not days.
AGV operation: fleet KPIs worth monitoring
An AGV system generates extensive logs and reports — the cheapest improvement tool you have, provided someone actually looks at the data. During operation, collect indicators and analyse them regularly, for example at a weekly meeting of the operations team. It is worth sticking to a few specific KPIs:
- Number of tasks per shift or per day — how many transport runs the fleet actually takes over from people.
- Percentage of tasks completed on time — whether the AGV delivers loads before they become a bottleneck.
- Average mission completion time — from the order to delivery of the load.
- Fleet utilisation — whether the vehicles spend more time driving or standing idle.
- Downtime and its causes — broken down into technical reasons and a lack of orders.
Monitor safety incidents separately: every emergency stop and every human intervention is worth breaking down to understand why it happened. The number of stoppages on its own says little — only the "causes" column shows what to improve. If AGVs frequently stop at obstacles, you probably need to enforce better housekeeping on the routes or improve zone marking. If one vehicle records noticeably more failures than the rest, check whether it is a hardware defect or the result of collisions with pallets set down incorrectly at the receiving station. Separating technical downtime from downtime caused by a lack of orders is the most important thing here: the former is a servicing problem, the latter a sign that the fleet has unused capacity.
Continuous improvement and scaling
Treat the AGV system as an element that evolves together with the factory. The KPI readings point directly to the way forward. When vehicles often stand idle, you can extend their scope of application with additional tasks and add missions, instead of keeping expensive equipment stationary. When, on the contrary, the system reaches its capacity limit because production has grown and the fleet can barely keep up, that is a clear signal to consider buying more units.
Ease of expansion is one of the real advantages of AGVs: adding vehicles or routes to meet new production requirements is usually simpler than with traditional, permanently installed conveying systems, and a well-designed layout can be moved to another location at little cost. Beyond scale, it is worth looking for savings in the configuration itself — optimising routes to shorten travel times and staying in touch with the supplier to learn about software updates or new modules that can bring additional benefits without replacing hardware.
Reviewing the results after six months
Roughly six months after start-up, carry out a fresh business assessment and compare the promises made during the analysis phase with reality. Compare current indicators — transport costs, cycle times, accident rates, the level of service to production — with those before implementation. If savings or productivity gains are visible, recalculate the ROI; the payback period may have shortened thanks to benefits that could not be foreseen at the start.
Share the results with management and with the crew. A concrete message — for example, "thanks to AGVs we have cut transport time by 20%, and accidents involving trucks have fallen to zero" — works better than any reassurances given before implementation. Highlighting success in numbers reinforces a positive attitude and genuinely makes it easier to accept further automation steps.
Retrospective and what next — closing the cycle
After the stabilisation period, hold a project retrospective: what went well, what could have been done differently, what to carry over to the next implementation. This is also the moment to identify further processes to automate — areas you deliberately set aside in the first phase which now, with a fleet and an experienced team in place, become a natural next step. The first implementation is usually the hardest to get going; each subsequent one builds on existing infrastructure and the knowledge you have gained along the way.
This completes the "step-by-step" methodology — from assessing the plant's readiness, through analysis, design, supplier selection, the pilot and training, to full operation. It is worth remembering, however, that AGV implementation is an ongoing process: the technology only delivers benefits when it is properly maintained and developed. A factory that treats operation as a separate, managed stage — rather than something "sorted once handed over" — will get far more out of automation.
AGV operation step by step — summary
The post-implementation phase can be reduced to a simple, repeatable routine:
- Make use of the supplier's start-up support, then set up a permanent reporting channel and monitor SLA compliance.
- Carry out regular inspections — sensors, batteries, wheels, markers — and keep a stock of spare parts.
- Train your own maintenance team to carry out minor servicing tasks.
- Track fleet KPIs: tasks per shift, percentage on time, average mission time, utilisation, and downtime with its causes.
- Separate technical downtime from a lack of orders — each calls for a different response.
- Add tasks when the fleet stands idle; buy more vehicles when it cannot keep up; optimise routes.
- After six months, assess the results, recalculate the ROI and present the results in numbers.
- Hold a retrospective and identify further processes to automate.
Well-managed operation turns a one-off implementation into a lasting advantage — it keeps the fleet running, provides hard data for decisions and lays the groundwork for further automation of the plant.
How often should AGVs be inspected?
In practice, quarterly inspections carried out by the service team are most common, but the final schedule depends on the supplier's recommendations and the intensity of the fleet's work. An inspection covers the technical condition of the vehicles (sensors, navigation, batteries, wheels), the navigation infrastructure and software updates. Regardless of the cycle, it is worth checking the basics daily — battery condition, sensor cleanliness — and keeping a stock of spare parts to shorten replacement time.
Which AGV fleet KPIs matter most?
The set depends on the goal of the implementation, but the core consists of: the number of tasks per shift, the percentage of tasks completed on time, the average mission completion time, fleet utilisation, and downtime together with its causes. The key is to separate technical downtime from downtime caused by a lack of orders — the former points to a servicing problem, the latter to unused capacity that can be loaded with additional tasks.
When should you buy more AGVs?
The decision is best based on operating data, not gut feeling. If the fleet utilisation indicator is rising and the system regularly reaches its capacity limit — for example, the fleet cannot keep up with the target number of missions at peak hours — that is a signal to expand. Before you buy more vehicles, check whether the problem can be solved more cheaply: by optimising routes to eliminate traffic conflicts or by changing the task schedule.
