How to scope an autonomous cleaning program for a large facility

Autonomous cleaning programs that disappoint were usually bought before they were scoped. A good scope for a shopping centre, airport terminal, hospital or warehouse turns the building into a set of zones, each with a floor type, a cleaning mode, a time window, a dock and a person responsible.

How much floor do you have to clean?

Start with cleanable floor area, not gross lettable area or building area. Walk the site with floor plans and mark out open floor that is cleaned on a repeating schedule: malls and food courts, terminal concourses and gate lounges, hospital corridors and lift lobbies, warehouse aisles and dispatch areas. Exclude tenancies, back-of-house rooms you do not clean, and anything cluttered with furniture that moves every day.

Then split that area by floor type, because the floor determines which machine you need. Record vinyl and linoleum, polished concrete, terrazzo, ceramic and porcelain tile (note deep grout lines), epoxy-coated warehouse slab, carpet and carpet tile, and entrance matting. Note transitions, thresholds, ramps, expansion joints and drainage grates, since these are where robots slow down, stop or need a wider turning path.

Which cleaning modes does each zone need?

Commercial cleaning robots generally offer some combination of the modes below. Many large sites need more than one, and the mix changes by zone.

Mode	Suits	Watch for
Dust mop	Polished concrete, vinyl, terrazzo between wet cleans	Does not remove sticky or wet soiling
Sweep	Warehouse slab, car park levels, loading areas with coarse debris	Dust generation near sensitive stock
Vacuum	Carpet in gate lounges, offices, hotel-style corridors	Loose cables, rugs and edging strips
Scrub (wet)	Hard floors in malls, concourses, corridors, food courts	Water use, drying time, slip risk while wet
Combined sweep and scrub	Large hard floors with mixed debris	Tank capacity sets the zone size

Wet scrubbing leaves a wet floor for a period after each pass. Safe Work Australia lists wet and polished floors among common slip hazards, so the scope should record when a zone is wet, how it is dried, and what signage or exclusion applies while people are nearby.

When can the robots clean?

Robots slow down and stop around people, so the cleaning window matters as much as the machine. One manufacturer’s deployment guide recommends running routes when an area is most free of people and freight, and moving floor care to quieter times if it is normally done while the facility is busy.

Each type of site has its own pattern. Shopping centres usually have a window after close and before trade. Airports run for most of the day, so scopes focus on overnight lulls and gaps between bank departures. Hospitals have corridors in use around the clock, which suits shorter, frequent routes rather than one long run. Warehouses clean between shifts or in aisles that are not being picked.

The same guide suggests breaking large spaces into smaller zones and keeping routes short enough to match the water tank, rather than teaching one long route across a whole floor. In practice, that means sizing each zone to what one tank and one battery charge can cover inside its window.

Where will robots dock, refill and drain?

This is where scopes most often fall short. For each zone, identify:

• a dock or parking spot out of public paths and egress routes
• a general power outlet within reach, and whether it is on a suitable circuit
• a clean water supply and a drain point for dirty water, or an auto-refill station if the platform supports one
• wireless or mobile network coverage at the dock and along routes, so alerts and reports get through
• lift access for multi-level sites, and whether the lift contractor will support integration

A cleaner’s room with a tap, a sink and a power point is often the natural base. If none exists near a large zone, plan a new service point rather than long daily transits.

Mapping, no-go zones and hazards

Robots are mapped by driving or guiding them through the site, usually when it is clear of people and temporary obstacles. Routes taught around objects that later move can leave gaps, so map during a quiet period with the floor in its normal state.

Mark no-go zones deliberately. Common examples are escalator landings and travelators, stair heads and open edges, glass balustrades and full-height glazing, loading docks, airside security boundaries, baggage reclaim belts, food court cooking areas, MRI suites, operating theatres and isolation rooms. Glass and mirrored surfaces can confuse some sensors, so flag them during mapping and test routes near them. Plan to remap after refits, tenancy changes or new fixtures.

How does it fit with the cleaning contractor?

Most large facilities outsource cleaning, and the program depends on that relationship. Agree up front who starts routes, who clears obstacles and responds to assist alerts, who empties and refills tanks, who reports damage, and who receives the coverage reports. Update the cleaning specification so it describes outcomes and frequencies by zone, not only labour hours.

At Christchurch Airport, contractor OCS New Zealand introduced an autonomous scrubber and a robotic vacuum alongside its cleaning team. OCS said the robots took on repetitive floor work so staff could concentrate on detailed cleaning and touch points, and one team member moved into a robot administrator role. Plan for the same arrangement, with robots on the contractor’s roster and named people accountable for them.

How should you measure coverage?

Before deployment, record a baseline: current audit scores, complaint logs and how often each zone is actually cleaned. After deployment, most platforms report area cleaned, run time, route completion and where the robot stopped or needed help. Some show planned against actual paths, which makes missed sections easy to spot and trace to obstacles, timing or route design.

Review these reports weekly at first, then monthly. Keep your visual audits running, because machine data shows where the robot went, not whether the floor meets your standard.

What do people still do?

Robots cover open floors on a schedule. People still handle spills and their signage, edges, corners and under seating, stairs, escalators, lifts, amenities, glass and detailing. In hospitals, the Australian Commission on Safety and Quality in Health Care notes that frequently touched surfaces such as door handles, bed rails and taps need cleaning more often than floors. Blood and body fluid spills need a trained person with the right products. Staff also prepare the floor, moving bins and signage and clearing obstacles before a route starts.

Scoping checklist

Item	What to capture	Why it matters
Cleanable area	Square metres by zone, excluding tenancies and cluttered rooms	Sets fleet size and route length
Floor types	Surface, finish, grout depth, transitions, ramps	Decides cleaning mode and machine type
Cleaning modes	Dust mop, sweep, vacuum, scrub per zone	Determines consumables and drying time
Cleaning windows	Hours per zone with low foot or vehicle traffic	Robots slow and stop around people
Docking points	Location, power, floor space, egress clearance	Avoids long transits and obstruction
Water and drain	Clean water supply, dirty water disposal, refill method	Tank capacity limits zone size
Network	Wi-Fi or mobile coverage at docks and on routes	Alerts, assists and reporting
Lifts and doors	Multi-level needs, lift contractor, automatic doors	Integration effort and approvals
No-go zones	Escalators, stairs, glass, docks, clinical and security areas	Safety and access control
Contractor roles	Who starts, assists, refills, reports and audits	Accountability after handover
Coverage measures	Baseline audits, report format, review cadence	Shows whether the program is working
Human task list	Spills, touch points, edges, amenities, detailing	Keeps standards where robots do not reach

Robotx can help you work through this scope on your own site before any equipment decision is made.