July 26, 2021—In mid-July, CCC Intelligent Solutions announced that it would add an artificial intelligence feature to its digital estimating platform, CCC One.
The company hopes to leverage its market share and expansive data trough to bring the AI technology to collision repairers. Mark Fincher, vice president of market solutions at CCC, told ADAPT that the technology has already been used in other sectors.
“We’ve had this AI technology deployed in the insurance market for some time now,” Fincher said. “This is the release for collision repairers that’s coming out at the end of August.”
In March, CCC said that the use of its AI among insurers had increased by 50 percent in a year for U.S. markets.
FenderBender reached out to learn a bit more about this technology for collision repair estimating.
Setting Up
The AI works from a set of photos taken shortly after a crash. Fincher says that these photos can be taken by the vehicle owner, if they’re able, and led by CCC’s interface. The basic process is to get all sides of the vehicles photographed, and the system will prompt the user for more detail if needed.
“We prompt them to take more detailed photos of the damage,” Fincher says. “And the AI is able to use those photos from that template to identify the damage quickly in a matter of seconds.”
There is a “heat map” feature that helps repair shops and vehicle owners visualize how the AI is working from the photographs to identify damage.
Shop Finalization
For shops, CCC and other developers of AI technology are marketing the perks of shorter cycle times and added efficiencies by letting the AI pick up some of the estimating work. That work is what Fincher calls “low-value tasks” that can be handled by the software.
“It’s going to allow estimators at repair facilities to focus on those more complex decisions,” Fincher says. “Leverage their expertise.”
Some human input is also needed to finalize AI estimates and make determinations on more heavily damaged vehicles. In some cases, the AI might hint at a potential issue that the human estimator needs to examine more closely for a final determination.
All that is to say that there are limits to AI estimates as the technology currently stands. Fincher says that with the historical data that CCC is able to input into the AI technology, it has gotten good at identifying the necessary elements for a lighter collision repair estimate. In the future, added technology might make the system more advanced.
For now, the goal is to find new efficiencies by pairing human and AI estimators.
“The two working together are going to give a more complete, accurate estimate in less amount of time,” Fincher says.
