An optional AI scan of the walk-around photographs that highlights areas worth a closer look before the driver signs the inspection off.
When a vehicle is collected, the driver has a few minutes in a yard or on a street to record its condition accurately. Miss a scuff on a bumper corner at collection and it becomes your scuff at delivery. Record something vaguely and it will not settle an argument weeks later.
TransportMate's AI Damage Scan is an optional step inside the inspection. The driver takes the walk-around photographs as normal, then chooses to run the scan; the photos are analysed together and any areas that look like damage are flagged against the specific body part, with a severity and a confidence level. The driver checks each flagged area at the vehicle and decides what goes into the record.
It is an assistant, not a judge. TransportMate does not claim the scan finds every defect, and nothing enters the inspection unless a person confirms it.
Almost every vehicle that moves has some pre-existing wear. The question a transport operator has to be able to answer is not whether the vehicle is perfect, but whether a particular mark was present before it was loaded.
Marks that were already on the vehicle need recording clearly at collection, or they become indistinguishable from transit damage.
Damage claims are usually decided on the quality of the evidence, not the strength of anyone's recollection.
Photographs taken at the vehicle, at the time of the handover, are the substance of the record.
The same structured walk-around at both ends means the two records can actually be compared.
The comparison between the two inspections is what identifies damage occurring in transit.
Each inspection is kept against its job with its photographs, time and location, and produces a PDF report.
The scan is intentionally cautious: it would rather raise a low-confidence flag for a driver to check than stay silent. That means some flags will be dirt, reflections or water rather than damage, which is exactly why a person verifies each one.
AI Damage Scan is available on plans that include AI features and requires a connection at the moment it runs. The rest of the inspection — photographs, damage markers, signatures — works offline as usual.
The walk-around captures the vehicle from the standard angles as part of the normal inspection flow.
AI Damage Scan is a deliberate, manual step. The inspection can be completed without it.
The set of walk-around images is assessed as a whole rather than one picture at a time.
Anything flagged is attached to a specific exterior part, with a short description, a severity and a confidence value, and the photo angles it was seen from.
Flagged areas are prompts to look closer. The driver confirms, adjusts or dismisses each one.
What the driver confirms becomes part of the inspection, with its damage markers and photographs.
The scan is one step in a chain that ends in a document you can send to a customer.
Captured in the app during the walk-around, attached to the job rather than sitting in a camera roll.
The structured set of panels and areas covered the same way at collection and delivery.
A point on the vehicle recording where the damage is, with its own supporting photographs.
Optional, and advisory: it suggests where to look and what it appears to be.
Only what the driver confirms, described and evidenced.
A PDF produced automatically on completion, containing the condition record and the photographs.
The inspection sits inside the collection or delivery proof, alongside the signature, time and location.
Generic inspection tools were mostly built around a checklist. Vehicle transport is different because the item being moved is itself the evidence, and it changes hands twice.
Multiple drivers, multiple standards. A consistent prompt to look closer helps level the quality of inspections.
Several vehicles loaded in quick succession, often under time pressure, where a missed mark at collection is expensive.
High movement volumes where condition evidence has to be produced reliably, not heroically.
Single-vehicle movements where the driver is alone and there is no second pair of eyes.
Part-exchanges and inter-site transfers where condition at handover has to be defensible.
Where vehicles regularly change hands between sites, drivers or suppliers.
Taken in the app at the vehicle, not uploaded later from elsewhere.
Recorded against the inspection itself rather than typed in afterwards.
The place the handover happened, captured with the record.
Signed on the device by the person releasing or receiving the vehicle.
A readable document summarising the inspection and its evidence.
Matching records at both ends of the movement, held against one job.
Compounds and car parks lose signal. Inspections must complete offline and sync later.
Something you can send a customer without assembling it by hand.
Damage evidence is not a separate system in TransportMate — it lives inside the collection and delivery records. A scan run during the collection inspection contributes to the Proof of Collection; the same inspection structure runs again at delivery and contributes to the Proof of Delivery. Because both sit on the same job, the two condition records can be read against each other, which is how a transit damage question is actually settled.
Inspections are one part of a job that continues after the vehicle is handed over. Once a job is completed with its POC and POD, TransportMate can carry it through to invoicing and on to your accounting system, so condition evidence and billing come out of the same workflow rather than two disconnected ones.
To be clear about what it does not do: detecting damage does not determine liability and does not change an invoice value. Damage findings are evidence for people to act on, not an automatic charge.
What is AI vehicle damage detection?
It is software that reviews photographs of a vehicle and highlights areas that appear to show damage, so an inspector can check and record them. It supports a human inspection rather than replacing it.
How does AI detect vehicle damage from photographs?
In TransportMate, the set of walk-around photographs is analysed together by a vision model, and anything that looks like damage is reported against a specific body part with a description, severity and confidence value, plus the photo angles it was visible from.
Can AI identify scratches and dents?
The scan reports common exterior damage types such as scratches, dents, chips, cracks and similar marks. It is not guaranteed to find every defect, which is why the driver verifies each flagged area at the vehicle before it is recorded.
What is the difference between AI damage detection and a manual vehicle inspection?
A manual inspection relies entirely on what the driver notices. AI damage detection adds a second pass over the same photographs that prompts the driver to look at specific areas. In TransportMate the manual inspection is still the record; the scan only suggests.
Can AI damage detection be used during vehicle collection?
Yes. It runs inside the collection inspection, after the walk-around photographs have been taken, and confirmed findings form part of the Proof of Collection.
Can AI damage detection be used during vehicle delivery?
Yes. The delivery inspection follows the same structure, so the scan can be run again and confirmed findings form part of the Proof of Delivery.
Can AI damage detection work with vehicle condition reports?
Yes. Confirmed damage, its markers and its photographs appear in the PDF condition report produced when the inspection is completed.
Can AI damage detection be used with POC and POD?
Yes — it sits inside the inspection that produces the POC and POD records, rather than being a separate process.
Does AI damage detection work offline?
No. The scan itself needs a connection at the moment it runs. The rest of the inspection, including photographs, damage markers and signatures, can be completed offline and syncs when the device is back online.
Run a collection through TransportMate, try the scan on the walk-around photographs, and judge the condition report it produces.