💡 Charge-Aware Routing How-To: Key Highlights
- Four inputs run the whole setup: orders, vehicle capacity/SOC, charger locations and delivery time windows are all a charge-aware routing engine needs to start proposing feasible routes.
- Charging time is a scheduling input, not an afterthought: a 17.7 kWh three-wheeler like the Bajaj Wego P9018 needs under 4 hours to reach 80% charge — that window has to sit inside the day’s route plan.
- Depot AC charging is the backbone, not the exception: globally, 91% of US and 59% of German electric truck charging happens overnight at depots rather than on the road (ICCT) — plan public fast charging as the backup case.
- Re-tag vehicle SOC and capacity weekly, not once at fleet onboarding — payload, battery degradation and traffic patterns all drift.
- When today’s orders exceed your EV capacity, mixed-fleet dispatch (holding back ICE or leased vehicles for overflow) beats over-promising an SLA you can’t keep.
If you read our piece on why AI-powered, charge-aware routing changes last-mile EV fleet economics, you already know the strategic case. This is the charge-aware routing how-to: the exact configuration steps an operations manager follows to go from a spreadsheet of delivery addresses to a route plan a driver can actually execute — without stranding a van at 40% state of charge three stops short of the depot. None of this requires new hardware. It requires getting four things into your routing system correctly: orders, vehicles, chargers and time windows, in that order.
Where Charge-Aware Routing Fits Into Your Day
Charge-aware routing sits between your order book and your driver app: it treats every vehicle’s remaining range and every charger’s location as a hard constraint, the same way it already treats a customer’s 2–4pm delivery slot. An AI operating system like YoMobility’s fleet management platform sits in that middle layer, holding your vehicle list, driver roster and charging data so route planning becomes one more consumer of data you’re already capturing. The five steps below are what you configure once, then tune weekly.
Step 1 — Import Orders and Depot Data
Start with the order file: pickup/delivery addresses, promised time windows, package weight and any fragile/priority flags, via CSV upload or an API feed from your order management platform. Alongside it, register your fixed points — depot address, operating hours, parking/charging bay count, and shift start times. Get this step wrong and every downstream step inherits the error: a route engine can only be as charge-aware as the order data feeding it is complete.
What good order data includes
- Delivery address with a resolvable lat/long — free-text addresses cause more failed-route calculations than any other input error
- A time window per stop (not just “today”) — even a loose 4-hour window lets the engine sequence charging around it
- Package weight/volume, since payload affects real-world range more than the rated figure does
Step 2 — Tag Vehicle Capacity, Battery Health and SOC
Every vehicle needs three tags before charge-aware routing can trust it: rated battery capacity, a real-world range derated for payload and traffic, and live state of charge from the vehicle’s telematics or OBD feed. Don’t route off manufacturer-rated range alone — a Bajaj Wego P9018 three-wheeler is rated for 296 km on its 17.7 kWh LFP pack, but a fully loaded cargo run in city traffic with AC and heavy stop-start driving will realistically return meaningfully less; tag the derated figure your fleet actually sees, not the brochure number. Mixed fleets should tag each vehicle class separately — a Tata Ace EV-class LCV with a ~20 kWh pack behaves very differently on a route than an 8–9 kWh cargo three-wheeler, even on the same delivery zone.
Static tags vs. live SOC
Capacity and derated range are set once and revisited monthly as batteries age. SOC is live — it should refresh every few minutes through the day so the route engine can re-plan if a driver runs hotter than expected (extra AC use, detours, idling in traffic). This is also where you connect vehicle management data — the same SOC and health telemetry your team already tracks for maintenance and uptime alerts becomes the routing engine’s live input, so you’re not maintaining two separate data sets for the same vehicle.
Step 3 — Mark Depot and Public Charging Points
Map every charger your fleet can reasonably use: depot AC points (with connector type and charge speed), any workplace or partner-site chargers, and public network chargers along common delivery corridors. Tag each with a realistic dwell time, not the fastest possible charge time — a driver stopping at a public charger to add 20 minutes of range mid-shift is a very different planning input than a van charging unattended overnight at the depot.
For most return-to-base delivery fleets, this step is simpler than it looks: research on electric truck and van charging in the US and Germany finds 91% and 59% of charging respectively still happens overnight at private depots, because delivery routes return to a fixed base every night. Build default routing logic around depot AC charging first, treating public DC fast-charging as the overflow option for high-mileage days. This is also where remote charging session management earns its keep — knowing which depot bays are occupied in real time is what lets the route engine trust a charger as “available” rather than just “installed.”
Step 4 — Let the System Propose Feasible Routes
With orders, vehicles and chargers all tagged, the system can solve for routes that respect every constraint at once: delivery windows, vehicle range remaining after payload and traffic, and charger dwell time where a mid-shift top-up is unavoidable. Set a minimum SOC reserve — most fleets run 15–20% — so the engine never proposes a route that lands a vehicle at the depot on fumes. Where a full charge genuinely can’t fit inside a shift (a 17.7 kWh pack needs roughly 4 hours to reach 80%), default that vehicle to overnight depot charging and build tomorrow’s route around a full battery, rather than squeezing a partial top-up into today’s plan.
Step 5 — Review, Override and Lock the Day’s Plan
Treat proposed routes as a first draft, not a final answer. A supervisor should review the plan each morning — check that priority/fragile orders landed on the right vehicle class, confirm no route runs tighter than your SOC reserve, and manually swap stops where local knowledge beats the algorithm. Once approved, lock the plan and push it to the driver app: locking stops the engine from silently re-optimizing mid-shift in ways a driver hasn’t seen, a common source of confusion in early rollouts.
Troubleshooting: When Demand Exceeds Your EV Capacity
Every fleet eventually hits a day where order volume outruns available EV range and charging windows — a festival surge, a depot charger down for maintenance, or simply more orders than usual. Charge-aware routing should surface this as a capacity gap before drivers are on the road, not after a van fails to make its last three stops.
Mixed-fleet overflow
If you run a mixed ICE + EV fleet, hold back a small ICE or leased-vehicle buffer for overflow days rather than running every vehicle at full utilization daily. Assign EVs to your densest, shortest-range zones first — where the cost and emissions advantage is largest — and route overflow to ICE vehicles on the longer, lower-density runs.
Stagger, don’t stack
Push flexible-window orders to a second wave later in the day, after the first wave of vehicles has had a partial recharge. This is usually cheaper than adding vehicles, and works well wherever same-day, not same-hour, is the actual customer promise.
Opportunistic top-ups
If a route already has a 15–20 minute dwell near a public charger (a hub delivery, a driver break), tag that charger as an opportunistic top-up point. It won’t fully charge a vehicle, but 15 minutes at a public DC charger can turn a route that was 10% short on range into one that’s feasible.
A Weekly Checklist to Keep Charge-Aware Routing Accurate
Configuration is a one-time setup; accuracy is a weekly habit. Run through this list every week so the system’s assumptions don’t quietly drift away from reality:
- Compare predicted end-of-route SOC against actual SOC for a sample of vehicles — a growing gap usually means derated range needs re-tagging
- Check charger uptime and any bays that went offline unreported
- Review how many routes needed manual override, and why — recurring overrides point at a bad default assumption, not bad drivers
- Confirm new vehicles, leased or owned, are tagged with capacity and SOC feeds before their first route
- Pull a route-completion and on-time-delivery report to catch any charge-related failures early, rather than at month-end
Most of this weekly review lives naturally inside fleet analytics — the same dashboard tracking cost per km and CO₂ avoided can flag SOC-prediction drift and route-override frequency without a separate spreadsheet.
Frequently Asked Questions
Four data sets: delivery orders with resolvable addresses and time windows, vehicle capacity and live SOC, mapped charger locations with realistic dwell times, and your depot’s operating hours. Most fleets already have three of the four in some form — the gap is usually live SOC feeds.
Plan for a full overnight depot charge (roughly 4–6 hours for a typical 3-wheeler cargo pack) rather than relying on mid-shift top-ups. Reserve public fast charging for opportunistic 15–20 minute top-ups during existing stops, not as a scheduled step in the plan.
Route the densest, shortest zones to EVs first, hold a small ICE or leased-vehicle buffer for overflow, and push flexible-window orders to a second wave later in the day. Charge-aware routing should flag the capacity gap the night before, not leave a driver to discover it mid-route.
For most return-to-base delivery fleets, depot AC charging should be the default — industry data shows the large majority of electric van and truck charging happens overnight at depots. Public fast charging is worth mapping as a backup for high-mileage or surge days, not as your primary plan.
Live SOC should refresh every few minutes automatically through telematics. Derated range and battery-health tags are worth reviewing monthly, and immediately after any new vehicle joins the fleet or an existing pack shows signs of degradation.
Sources: ICCT — Charging Solutions for Battery-Electric Trucks (2022) | RMI — Clean Energy 101: Electric Trucks | Bajaj Auto — WEGO P9018 Specifications
Set Up Charge-Aware Routing For Your Fleet
Tell us your fleet size and delivery zones — we’ll show you how YoMobility configures charge-aware routing on your vehicles.