Automation
A practical payback analysis for AGV and AMR projects, including the cost drivers most business cases leave out.
Last updated: 11 October 2026
Automation is sold on labour cost. That is the smallest number in the calculation. Here is what actually decides whether a project pays back.
AGV vs AMR — the real difference
- **AGV** follows a fixed path — magnetic tape, QR codes or reflector markers. Cheap, proven, and fast to deploy. Change the layout and you re-lay the tape.
- **AMR** navigates by SLAM, building a map of the building. No infrastructure. Change the layout and you redraw a map on a laptop.
The cost drivers nobody puts in the business case
| Item | What it actually costs |
|---|
| Infrastructure | AGV needs tape or markers laid — and re-laid every layout change |
| Downtime during rollout | 2 – 6 weeks of disrupted picking while the fleet is commissioned |
| WMS integration | Often 30 – 50% of the project cost, and the usual cause of overrun |
| Traffic management | Software to stop robots deadlocking — not optional above about 5 units |
| Battery and charging | Opportunity charging stations, or a swap routine |
| Maintenance skills | Someone on site has to be able to change a wheel and read a fault log |
| Change management | Operators who do not trust the robots will route around them |
What actually justifies it
- **Night shifts you cannot staff.** The strongest case. A robot does not need a second shift premium.
- **Repetitive point-to-point moves** over 50 m, more than 100 times a shift
- **Accuracy requirements** — barcode scan at every pick, no exceptions
- **Hazardous or refrigerated environments** where people should not work
- **Growth you cannot hire for** in a tight labour market
What does not
- Low, variable volumes — the robot sits idle and the payback never arrives
- A layout that changes every quarter
- A warehouse where the actual bottleneck is the goods-in process, not the transport
- Any site that has not first fixed its racking layout and labelling
A realistic payback
Typical AGV or AMR projects in warehousing pay back in 2 to 4 years where the labour being replaced is genuinely unavailable — second shifts, weekend shifts, cold stores. Where it is simply replacing labour you already have, payback stretches past five years and the project is usually a technology purchase rather than an investment.
Need help choosing the right truck?
Send us your load weight, lifting height and aisle width — we will recommend the right model.
Get a Quote
Frequently asked questions
How much more does an AGV cost than manual labour?
Capital cost is typically one to three times a year of manual labour, but over five years and two shifts the total usually comes out lower.
When does automation not pay back?
Low move counts (under 20 a day), frequently changing routes and poor floors all extend payback considerably.
What is a typical AGV payback period?
Two to three years for two-shift operations above 50 moves a day. Single-shift, low-frequency sites may exceed five years.