Industry
AI for Vehicle Recycling
Pricing engines, fleet management, and compliance automation for UK Authorised Treatment Facilities. Built by someone who has worked in vehicle recycling for 20 years.
The Problem
Most UK vehicle recyclers still price cars the same way they did ten years ago. A customer phones in. Someone looks up the registration, checks the scrap metal price, estimates the kerb weight, factors in the collection distance, and gives a quote. That process takes 10 to 15 minutes per vehicle. On a busy day, an operator might handle 30 or 40 calls. Missed calls mean lost vehicles to competitors who answer faster.
Pricing is only one part of it. Collection scheduling runs on phone calls and diary entries. Drivers confirm pickups by text. The office has no live view of where the fleet is or which jobs are done. When something goes wrong — a vehicle is not where the customer said, a driver is running late — the office finds out by phone call, usually from an unhappy customer.
Then there is compliance. Certificates of Destruction must be issued to the DVLA. Depollution records must be maintained for every vehicle. Quarterly returns go to the Environment Agency. Hazardous waste consignment notes need to be accurate. Most ATFs handle all of this manually — spreadsheets, paper forms, and a few days of intense work at the end of each quarter. A single error can trigger an audit.
These are not technology problems in the abstract. They are operational problems that cost real money: lost customers, wasted driver hours, compliance fines, and the constant risk of the one person who knows the pricing spreadsheet being unavailable.
The Solution
AI addresses these problems at three levels: pricing, fleet operations, and compliance. Each can be deployed independently or as part of a connected system.
AI Pricing Engine
A customer enters their registration number on a web portal. The system looks up the vehicle via DVLA data, calculates a price using live scrap metal rates and the vehicle's kerb weight, applies zone-based collection costs from the customer's postcode, and returns a quote in under 30 seconds. The operator sets the pricing rules — minimum margins, premiums for catalytic converters, deductions for non-runners. Every quote is logged with full context.
The result: consistent pricing regardless of who is on shift. No more guesswork. No more missed calls while someone manually prices a vehicle. And a full record of every quote — including the ones that did not convert — so you can see where you are losing business and why.
Fleet Management
A driver app — a progressive web app that runs on any smartphone — replaces the phone-and-diary system. Jobs are assigned with a tap. Drivers see their route ordered by postcode proximity. They mark jobs as collected, flag issues, and upload photos on collection. The office sees live status without making a single phone call.
Route optimisation groups collections by area, suggests the most fuel-efficient order, and flags when a driver is near a pending collection. Over time, the system learns which postcodes have access issues and which time slots have high no-show rates.
Compliance Automation
When a vehicle is processed, the system pre-fills Certificate of Destruction details from the purchase record, generates depollution checklists based on vehicle type, and compiles quarterly data automatically. Proactive alerts flag when a vehicle has been on site too long without a CoD, when depollution records are incomplete, or when a reporting deadline is approaching.
Case Study: 15-Minute Quotes to 30-Second Quotes
At Reclamet Recycling, pricing a scrap vehicle used to take 10 to 15 minutes per call. An experienced buyer would look up the registration, check the scrap price, estimate kerb weight, calculate collection distance, and adjust for the state of the car. Pricing was inconsistent — two people quoting the same vehicle on the same day could be £30 to £50 apart.
The AI pricing engine changed that. Registration lookup, weight-based pricing from live metal rates, zone-based collection costs, and operator-defined margin rules — all calculated in under 30 seconds. Pricing became consistent. The quote log revealed that 40% of lost customers were in postcodes where collection costs were too high. Zone boundaries were adjusted and volume recovered.
The system is built on React with Supabase. Vehicle data comes from a registration lookup API. Metal prices update from published indices. Pricing logic runs server-side so rules are not exposed to customers. Every quote is logged with full context — vehicle details, price offered, influencing factors, and conversion outcome.
Pricing
| Module | Upfront | Timeline | Ongoing |
|---|---|---|---|
| Pricing Engine | £3,000 – £8,000 | 2–4 weeks | £200 – £400/month |
| Fleet Management | £5,000 – £12,000 | 4–6 weeks | £300 – £500/month |
| Compliance Automation | £4,000 – £10,000 | 3–5 weeks | £200 – £400/month |
All projects start with a working prototype in the first 2 weeks. You see real output before committing to a full build. Ongoing costs cover hosting, support, and API usage. You own the code and the data.
See the pricing engine in action
30-minute call. I will walk through the system, answer your questions, and tell you honestly whether AI fits your operation.
Book a Free Discovery Call →AI developer and founder of NikkSi. 20 years of commercial operations experience at Reclamet Recycling, covering sales, pricing, fleet management, and compliance. Four AI products shipped.