Fleet electrification maths gets done on vehicles and chargers — range, payload, kW, unit cost. Then the vehicles arrive, and the depot discovers that its real problems are software problems: will every vehicle be ready at 6am, on a grid connection half the size of the theoretical peak, and can anyone prove what charging actually costs per route?
Public charging is a retail business; depot charging is an operations business. Different physics, different KPIs, different software requirements. Here's the checklist we'd walk any fleet through before the first charger is ordered.
1. Readiness, not speed
The depot's defining metric isn't charge speed — it's departure readiness: every vehicle at its required state of charge when its shift starts. A van that finishes charging at 2am gained nothing over one that finishes at 5:45am; a van at 60% at 6am is a failed route.
That reframing changes the software requirement completely. You need charging scheduled against departure times — vehicles mapped to duties, energy needs computed per vehicle, charging sequenced across the night to hit deadlines within the site's power budget. Speed is a resource the scheduler spends, not a goal.
Ask your platform: can charging be orchestrated toward per-vehicle readiness targets — and what happens to the plan when a vehicle plugs in late?
2. The power budget is the real constraint
Almost no depot gets a grid connection sized for every charger at full power — nor should it pay for one. As we covered in how load management actually works, static and dynamic sharing let a modest connection serve a large fleet, because overnight dwell time is enormous compared to charging need.
The depot-specific wrinkle is priority. Even sharing is the wrong policy when the 4am refrigerated run matters more than the 9am admin pool car. Priorities are operational policy, and they change — which is why they belong in a rules layer your ops team can read and edit, not in charger firmware.
Ask your platform: show me a site power budget, dynamic headroom from the site meter, and a priority rule — live.
3. Failure tonight is a route lost tomorrow
Public charging failure costs a session; depot charging failure costs tomorrow's delivery schedule. A charger that faults at 11pm with a vehicle attached must be someone's problem at 11:05pm — an alert, a suggested action, and ideally an automatic re-plan that shifts the vehicle's energy need onto healthy chargers within the budget.
This is where outcome-based monitoring and predictive maintenance stop being uptime hygiene and become schedule protection: the connector flagged for rising failure risk gets fixed on Tuesday afternoon because Wednesday's routes depend on it.
Ask your platform: when a charger fails mid-charge overnight, what happens — in the software, before any human wakes up?
4. Energy cost is now a route cost
Diesel had one price per litre; electricity has a price per half-hour. Charging the fleet between the right hours can change the energy line dramatically — but only if the scheduler knows the tariff and the CFO can see the result.
The reporting requirement is specific: kWh and cost per vehicle, per route, per depot, over time — reconciled against readiness. That's the dataset that proves (or disproves) the electrification business case, feeds mileage reimbursement where drivers home-charge, and eventually supports tariff and flexibility decisions. It needs to flow into fleet management and finance systems via API, not live in a portal screenshot.
Ask your platform: export cost per vehicle per week, and show me how it reaches our fleet management system.
5. Drivers and depot staff are users too
Depot charging has an underrated human layer: the driver returning at 1am who needs to know which bay, the shunter re-plugging vehicles overnight, the transport office needing tomorrow's readiness board at a glance. If the software serves only head office, plug-in compliance becomes the nightly failure mode — and no scheduler can charge a vehicle nobody connected.
Simple, role-appropriate surfaces — a readiness board, plug-in confirmations, exception alerts for "vehicle expected but not plugged in" — do more for depot reliability than another kW of charger. (Role-based access matters here too: the shunter needs the plug board, not the tariff editor.)
Ask your platform: what does the 1am driver see? What does the transport office see at 5am?
6. Count on mixed hardware from day one
Fleets buy chargers in waves, across brands, as vehicles arrive — and depot hardware lives a hard life, so replacements accumulate variety. Everything we've written about mixed-fleet OCPP support applies doubly at the depot: the scheduler must treat a five-year-old 7 kW post and a new 150 kW rapid as one orchestrated pool.
Ask your platform: which hardware brands and OCPP versions are running in depot deployments today?
The one-line summary
Buy chargers for the electrical job; choose software for the operational one. A depot that gets readiness scheduling, power budgeting, overnight failure handling, cost reporting and human surfaces right will outperform a depot with faster chargers and none of the above — every single morning.
AmpNexus runs depot charging on the same platform as public networks — readiness-aware smart charging, site budgets, predictive maintenance and per-vehicle reporting. See Smart Charging or talk to us.