TLDR: Damage AI looks at the photos your team already takes during an appraisal, finds the scratches, dents, chips and kerb rash, classifies each one and attaches a repair cost estimate. The result: reconditioning is priced before you buy the car, every deduction has photographic evidence behind it, and the condition standard is identical across every appraiser, shift and site. No special hardware — just a smartphone.
Ask any used-car manager where gross goes to die, and reconditioning surprises will be in the first three answers.
The car looked fine in the lane. The offer went in. Then it hit the workshop and the “tidy little trade” needed two panels, a wheel refurb and a windscreen chip repair that someone should have seen. That money comes straight out of the deal — after the deal is done.
Damage AI exists to kill that moment. Here’s what it actually is and how it works.
What Damage AI does
Damage AI is image-based damage detection with recon pricing built in. During the normal appraisal or inspection walk-around, your team photographs the car — guided capture makes sure the angles and panels are covered. The AI then does three things with those photos:
- Finds the damage — scratches, dents, chips, kerb rash, marked directly on the image, including the ones a quick walk-around misses.
- Classifies it — what kind of defect, on which panel, how severe.
- Costs it — each defect gets a repair cost estimate, and the estimates roll up into a total reconditioning figure for the car.
The output isn’t a red flag that says “damage somewhere”. It’s an itemised, photo-backed recon bill, produced in the time it takes to walk around the car.
No rigs, no tunnels, no new hardware
A fair question: doesn’t this need special equipment?
No. Damage AI works from ordinary smartphone photos. There’s no scanning tunnel to install, no rig in the drive lane, no hardware to maintain. The guided photo workflow in EvalExpert and InspectExpert is the only “equipment” — it exists to make sure every car gets the same angles, so the AI (and any human reviewing later) is always working from complete evidence.
That matters for adoption. Appraisers will follow a photo guide on the phone they already hold. They will not queue cars through a tunnel.
Why consistency beats the sharpest eye
Your best appraiser is genuinely good at spotting damage. But your best appraiser isn’t on every appraisal — and even they see cars differently at 9am on a quiet Tuesday versus 5pm on a slammed Saturday.
The AI applies one standard, every time, everywhere. Same detection threshold on car one and car forty. Same standard at the flagship site and the satellite lot. For dealer groups, that consistency is the whole game: group management can finally trust that “condition-adjusted” means the same thing across every branch. It’s the same logic that makes centralised trade-in insight possible — consistent inputs, comparable outputs.
And because every finding is marked on the photo, a human can always review, override or add context. The AI does the tedious part; your people make the call.
What it changes at the moment of offer
Two things, both worth money:
The deduction becomes defensible. “We’re allowing $1,850 for reconditioning — here are the four items, photographed and costed” is a conversation. “We knocked two grand off for condition” is an argument. Customers accept itemised evidence far more readily than round numbers, which means the offer holds without you buying the car twice.
The gross is protected before you own the car. The recon estimate lands in the appraisal, so it’s reflected in what you pay — not discovered in the workshop a week after the trade. That’s the difference between pricing risk and absorbing it. It’s step two of the process we laid out in our trade-in valuation guide: cost the damage, don’t guess it.
Beyond the trade-in lane
The same detection and costing runs anywhere condition needs documenting:
- Fleet de-fleet — process outgoing units at scale with one consistent standard, so you know the condition of every car before it hits wholesale.
- End-of-lease returns — give lessees an itemised, photo-backed bill that stands up when it’s challenged.
- Auction and wholesale listings — a costed condition report attached to the listing means buyers bid with confidence, and fewer sales come back to bite.
- In-stocking checks — catch transport damage the day the car arrives, not the day it’s being detailed for a customer.
The flywheel behind it
One more thing worth knowing: detection and costing models improve with volume. Damage AI learns from the appraisals, inspections and repair outcomes flowing through 1,200+ showrooms in 14 countries — so the estimates track what reconditioning actually costs in the market, not a static rate card. Every inspection makes the next one sharper.
If reconditioning surprises are still finding their way into your deals, book a demo and run Damage AI over a few of your own trades — the walk-around takes about the same time it does today. The surprises just stop being surprises.