Despite advances in AI-powered claims processing, human judgment “remains key,” says Mitchell – Repairer Driven News | CarTailz

Automating the auto claims process is “not as complete as the industry once thought,” and “manual assessments and human judgment remain key to the estimation process,” says Mitchell’s parent company, Enlyte, in its 2022 annual trends report.

Full automation should be considered a “long-term goal” as AI-powered claims processing strives to live up to the hype that has been built around it, says Olivier Baudoux, Mitchell’s senior vice president of global product strategy and artificial intelligence in Chapter 4 of the report titled Claims Automation: Separating Fact From Fiction.

Jack Rozint, Mitchell’s senior vice president of repair sales, told Repairer Driven News that Mitchell believes “very strongly” in the use of AI in the estimating process, which he described as “very workable and very promising.”

The message for those disappointed that the technology isn’t meeting unrealistic expectations is: “Hang on – it’s going well, and there’s a lot more good stuff to come.”

While the pandemic has accelerated advances in straight-through processing [STP]”There is still work to be done to deliver on the promise of a no-touch entitlement,” the report said. “As an industry, we are just beginning to reconcile initial expectations with the current state of the art.”

The report considers changing expectations of technology through the Gartner Hype Cycle, a graphical presentation developed, used and branded by research firm Gartner to represent the maturity, acceptance and social application of specific technologies. On this curve, the report says, expectations lie between Stage 3, disenchantment, and Stage 5, mainstream acceptance.

“The key to this phase is recognizing what is possible and how technology can be used in situations that will bring the most benefit,” the report says. Two areas where this is true are estimating and total loss claim triage and automation.

The report appears to support concerns that have long been voiced by the accident repair industry. Addressing the issue of photo-based estimation in April 2021, Wired magazine found a common complaint among stores: damage identifiable by a personal human appraiser is overlooked in photo estimation.

“While STP is a fundamental goal and certainly the long-term vision, it is equally important to recognize that manual assessments and human judgment remain key to the estimation process,” the report concludes. “In the case of major accidents or damage involving the latest vehicles, the involvement of experts still brings tangible business benefits,
such as preventing cost losses. Therefore, optimizing the core calculation solution to improve the user experience remains a priority. This may include showing up the right photos at the right time, hiring an appraiser if the part chosen isn’t clear, or even asking a “real” person to confirm what the machine may not be sure about.”

The report finds that while 97% of surveyed auto insurers recognize the value of STP, only 10% are making significant use of the technology and fewer than 25% of global organizations have developed an enterprise-wide AI strategy.

As AI’s shortcomings became apparent, “claims automation sentiment went from hopeful to hopeless,” the report says, with many companies scaling back their STP investments from operational goals to research and development projects. “They lost faith in non-contact estimation and instead started focusing on simpler and more realistic use cases for AI — like

LexisNexis Risk Solutions, which not long ago reported that 79% of auto insurance companies were open to the idea of ​​touchless claims, now predicts that telematics data and AI will result in 60% of claims being checked through automation by 2025, it said in the report.

“[W]With only half of non-injury claims expected to be fully automated by 2025, this is now considered a long-term goal,” it said.

Mitchell’s report acknowledges some of the ways in which AI-assisted estimation has fallen short, but notes that progress is being made in a number of areas.

For example: “If you can’t decipher a vehicle identification number (VIN) and uniquely identify the model’s options, how can you accurately predict which parts need repair or replacement?” Do you use a bumper with sensors? What about headlights? And if the part is made of a special material like aluminum, how should the repair schedule differ?”

Although the AI ​​was able to recognize the “key” vehicle components, it became clear that “it had hundreds or even thousands more to learn,” the report said.

The AI ​​also lacked the ability of human assessors to review all available photos in context before making a decision, or to analyze multiple photos to determine unrelated past damage. On the other hand, human appraisers are “unwilling to embrace their new role in reviewing and approving pre-written estimates generated by AI,” the report states.

Experience has shown that “while automation is designed to facilitate claims settlement, it has clearly failed to replace human assessors,” the report states.

Mitchell claims progress has been made in a number of areas. Mitchell’s proprietary AI solution, Intelligent Damage Analysis (MIDA), can now predict more than 300 internal and external parts and make recommendations for operations such as repair and replacement, remove and install, and rework and merge. These “align with OEM repair procedures,” confirmed Rozint.

MIDA is also able to largely decode VINs by combining vehicle information with AI. “Vehicle, repair and historical damage data are critical for a non-contact estimate to create an accurate appraisal,” it says.

Recycled and aftermarket parts have been integrated into the system and the AI ​​is now able to correctly identify the primary point of impact as well as distinguishing any unrelated prior damage.

The report suggests that STP technology will be measured by profits made on certain specific metrics. These include the number of predicted guess lines that are correct based on what is considered “ground truth”, the number of parts successfully matched; the number of incorrectly added calculation lines to be revised and the number of calculation lines are missing or incorrect and require manual intervention.

In these metrics, MIDA has shown an increase of 16% in the percentage of estimate lines correctly auto-completed, 26% in parts properly recognized, 12% in correctly identifying damaged parts, and 8% in repair and replacement operations accounted for.

More information

Wired: AI is coming to the auto shop, and body shop owners aren’t happy


Featured image by Kwangmoozaa/iStock.

Share this:

Leave a Comment