Service Robot Fleet Energy Budget and Route Optimization

The Energy Budget: Where the Day Actually Goes

Start with a number most datasheets omit: the energy a unit consumes per hour of real work on your floor, not on a test track. It is always higher than the nominal figure, because real floors add friction from carpet, ramps, frequent stops and door and lift interactions. Build the budget from four components and measure them during the pilot rather than trusting the brochure.

Component	What drives it	How to measure it
Traction	Floor type, gradient, payload, distance per cycle	Amp-hours consumed per measured cycle during the pilot
Cleaning or delivery work	Brush or pump load, water draw, door mechanism cycles	Consumption with the work head engaged versus idle
Compute and sensing	Navigation, lidar, cameras, always-on connectivity	Baseline draw while parked but powered, then working draw minus baseline
Auxiliary	Heating, cooling, lights, screens, lifting	Measured separately for any unit that runs an auxiliary load

The working figure is amp-hours per hour on your floor. Multiply by shift length to get the energy a full shift demands, then divide by usable battery capacity, accounting for the depth-of-discharge limit the cell can safely take, and you have the number of charges a single unit needs per shift. If that number exceeds one, a single unit cannot hold a full shift, and the fleet size or the charging plan must change. The fleet-sizing logic this feeds into is the same arithmetic described in the throughput planning guide, with energy rather than coverage as the binding constraint.

The Charge-to-Work Ratio and Fleet Sizing

The number that decides whether a fleet is viable is the ratio of charging time to working time. A unit that works four hours and charges two has a duty cycle of two thirds; a unit that works four and charges one has four fifths. That ratio, not the headline battery capacity, sets how many units are needed to cover a scene across a shift.

Work it through for a concrete case. A scene needs 6 productive hours of coverage per day. A unit with a two-thirds duty cycle delivers 4 hours of work per 6-hour block, so it covers two thirds of the requirement and the remaining third needs either a second unit or an opportunity charge. Choosing between those options is a cost decision: a second unit carries capital cost, while opportunity charging carries schedule cost and requires the unit to be free at the right moment. The economics of that trade-off sit alongside the wider cost model in the total cost of ownership ledger.

Charge Rate: What the Dock Actually Delivers

Charger recovery rate is routinely overstated by the assumption of a full-rate charge across the whole cycle. In practice the rate tapers as the cell fills, so the last portion of a charge takes disproportionately long. Plan on the usable window, not the peak figure, and distinguish two strategies.

Strategy	How it works	Best used when
Full-cycle charging	Return to dock at a low state of charge, charge to full, resume	Long blocks of continuous work and a generous off-shift window
Opportunity charging	Top up during natural pauses, never deeply discharged	Frequent short tasks, multiple docks, dense but interrupted work

Opportunity charging keeps the cell in its most efficient mid-range and avoids the slow tail of a full charge, but it depends on docks being where the robot already is. That placement decision belongs to the survey, not to the electrician who arrives after commissioning. Plan the dock positions against the work pattern so a top-up costs no travel, and confirm the electrical supply at each position during the readiness review described in the facility readiness assessment.

Route and Sequence: Cutting Energy Per Square Metre

Routing affects energy more than most operators expect, because every unnecessary turn, reverse and re-visit is traction energy spent twice. Three choices consistently reduce consumption per square metre without changing hardware.

• Boustrophedon over adaptive: on open floors, straight parallel passes with a single turn at each end beat adaptive wandering. Reserve adaptive behaviour for obstacle-dense areas where a straight line is impossible; do not let the default mode run the whole floor.
• Zone the scene to the unit: assign contiguous zones rather than interleaving, so a unit finishes where it is rather than travelling across the building between partial areas. Fewer long transits is less energy for the same coverage.
• Sequence heavy work early: run the highest-draw task first, while the battery is in its strongest state, and leave light or low-draw zones for later in the block. This does not reduce total energy but it delays the point at which a mid-shift charge becomes unavoidable.

Measure the effect of each change as amp-hours per square metre, not as a judgement, and hold the sequence that wins. A route change that cuts energy per square metre by ten percent extends the shift by roughly the same proportion at no hardware cost. Pair this with the productivity measurement described in the cleaning productivity baseline method so that route changes are judged against a real pre-change figure.

Scheduling: Moving Charging Out of the Working Day

The most effective energy measure does not cut consumption at all, it moves charging to a time when the robot would otherwise be idle. If the building is occupied and cleaned during the day, off-hours charging removes the dock from the critical path entirely and the duty-cycle penalty disappears. The scheduling discipline for that arrangement is the same as any night-shift plan, described in the night shift scheduling guide.

Where charging cannot be moved off-shift, stagger the docks so units return at different times. Three docks all approached at once produce a queue that costs more lost work than the charging itself. Stagger by zone completion order, and reserve one dock as a fast top-up point for whatever unit is furthest behind. Finally, track energy in the monthly reporting alongside the utilisation and coverage numbers so a slow drift, whether from a degrading battery or a changed floor, surfaces as a trend rather than as a mid-shift surprise.

Frequently Asked Questions

Should I buy a second battery per unit? Only if the unit is designed for hot-swap and the operation can absorb the swap pause. For most fleets, opportunity charging at well-placed docks is cheaper and simpler than a battery pool, which adds handling, storage and safety obligations.

How do I know if a battery is degrading? Watch amp-hours delivered per charge against the same route over months, not the percentage readout. A steady fall in delivered energy at a full charge is the signal, and it typically precedes any visible runtime complaint.

Can I run a unit down to a low state of charge to extend the block? Repeated deep discharge shortens cell life, and the runtime gained late in the block is small. Keep within the depth-of-discharge limit the cell is specified for and plan the charge, rather than borrowing hours from battery lifespan.

Does fast charging solve it? Only if the docks and the supply can sustain the rate and the cell accepts it without accelerated ageing. Treat fast charging as a tool for the fast top-up dock, not as a substitute for a charging schedule.