Artificial intelligence is starting to move beyond fleet dashboards and reporting, with DingGo developing technology designed to turn photographs of vehicle damage into information fleet managers can use to make repair decisions.
Speaking during the supplier panel at Everything Electric Sydney, Dmitri Feigin from DingGo explained how the company’s Crash Intelligence technology is using unstructured information such as damage photographs and repair quotes.
DingGo joined FleetGuru, Geosecure and Origin Energy on the panel as supplier sponsors of the Fleet News Group Awards program.
For fleet managers, the potential benefit isn’t simply a faster damage assessment. It’s being able to make an earlier decision about what happens to the vehicle.
Turning photos into fleet data
Fleet management systems have become increasingly effective at analysing structured information such as kilometres travelled, fuel consumption, maintenance expenditure and repair costs.
Damage photographs present a more difficult problem.
Feigin said DingGo had been working on extracting useful information from that unstructured data.
“The harder problem to solve, and the harder unlock that we’ve been working on for a couple of years at DingGo now, is actually using all the unstructured data, which is the photos of that damage, the repair quotes, and starting to actually extract value from that,” he said.
DingGo’s Crash Intelligence technology is designed to analyse photographs and develop an assessment of the damage.
An assessment in seconds
Feigin said the technology could move from photographs to a comprehensive damage assessment in less than 30 seconds.
The next question is what the damage is likely to cost. For each damaged area, the system considers whether a component is likely to require replacement or could potentially be repaired.
It can also assess whether the damage may suit alternatives such as mobile repair, paintless dent repair or detailing.
“Then based on that, give you a cost estimate, so you can go from photos to a really good high-level quote that’s quite accurate in under 60 seconds,” Feigin said.
The system also considers the likelihood that the vehicle could be a total loss.
Repair now or keep the vehicle working?
For fleet managers, speed creates more options.
Traditionally, a damaged vehicle may need to travel through several stages of inspection and quoting before the fleet manager has enough information to decide what to do.
AI-based assessment potentially moves that decision much closer to the time of the incident.
“Do I repair now? Do I repair later and keep a car on the road for a bit longer because it’s drivable?” Feigin said.
“Actually, do I maybe not repair at all and sell the car with pre-existing damage?”
Those questions are particularly relevant to large fleets where relatively minor damage doesn’t necessarily mean a vehicle needs to be immediately removed from service.
AI becomes a fleet decision-making tool
The development also illustrates a broader change in the role of artificial intelligence in fleet management.
Instead of simply producing another report, AI tools are beginning to interpret information that previously required human expertise and convert it into data that can support an operational decision.
That could become increasingly useful as Australian fleets add vehicles from new brands and adopt different drivetrains.
Historical experience may be limited when an unfamiliar vehicle is damaged. Large datasets can potentially help fill some of that knowledge gap.
For fleet managers, the promise of AI isn’t that the technology makes the final decision. It’s that better information could reach the person making that decision much faster.
