Fixed Ops Journal - February 2018 - 38

FIXED OPS JOURNAL

LEARNING
MACHINES
How artificial intelligence is already disrupting fixed operations



MARK ELIAS

So-called smart parts can send data about
use and wear directly to the software, Brooks
says. The AI systems not only predict parts
ary Brooks scheduled a service apfailures, he says, but help automakers price
pointment last year for a tire recall
replacement parts more accurately by analyzon his car. After he waited at the
ing sales trends.
dealership for 45 minutes, a service
Mazda North American Operations is using
adviser told him the parts department didn't
AI to help transform its parts distribution syshave the needed tires in stock.
tem from managing a network based on re"I said, 'How can this be?' " Brooks recalls.
gional centers to creating a global supply
"It's not as though I was a walk-in customer."
chain, says Robert Davis, Mazda's senior vice
Brooks is doing more than complaining
president for special assignments.
about parts shortages at dealerships. He works
"Our wish list is to have the technology to be
for Syncron, a software developer that uses artiable to predict a parts failure on a car, comficial intelligence to create products designed
municate it to the dealer, to the parts distributo help automakers and dealers improve suption center, all the way back to the core suppliply processes for replacement parts. Brooks is
er," Davis told Fixed Ops Journal.
the company's chief marketing officer.
Such an AI-based system, Davis
Car companies, automakers, dealsays, also will notify the vehicle owners, insurers and suppliers are maker: "It looks like a part will be failing
ing big investments in AI - essentialon your car within the next 3,000
ly, the simulation of human intellimiles, so why not bring it in so we can
gence and learning by machines - to
fix it? The part will be here."
make fixed operations more efficient.
Repair order fill rate is a better meaAmong the areas to which AI is being
sure than overall fill rate, Davis says. "If
applied: parts supply chains, collision
your car needs three parts, and I only
assessments and repair processes.
Davis: AI aids in have two in stock, those two parts become completely irrelevant," he says.
Predicting parts
parts supply.
In their work with automakers, Syncron and its competitors use AI to predict the
Better crash appraisals
parts that dealership service departments are
AI also is disrupting the way insurers exammost likely to need, and to provide them
ine cars and trucks damaged in crashes, estipromptly.
mate parts and labor costs for repairs and
AI enables dealerships to save money by
send the vehicles to body shops, including
maintaining leaner parts inventories and orthose operated by dealerships.
dering fewer parts that are likely to be reTractable, a British AI firm, is working with
turned to the automaker or become obsolete,
Mitchell International, a developer of autoBrooks says.
motive software in San Diego, to speed acciSoftware systems monitor dealerships'
dent repairs.
parts orders to automakers. Algorithms anaTractable uses a database of 300 million imlyze supply and demand patterns to "learn"
ages of damaged vehicles to help appraise
about parts runs in specific regions.
photos of crashed cars and determine needed
Brooks says Syncron's software has helped
repairs. The company says that computers ussome automakers increase their fill rate - the
ing the algorithms in its AI Review software asshare of customer parts demand that is met
sess damages better than human estimators.
through immediate dealership stock availAI-based programs improve the prospects
ability - from 60 to 90 percent.
of "a proper and safe repair," says Olivier Bau-

PAGE 38

foj@autonews.com

FEBRUARY 2018

AI in the service bay
How dealership fixed operations benefit
from artificial intelligence
 Computers predict the replacement
parts dealerships are likely to need
to order from automakers.
 Software analyzes hundreds of
millions of images of damaged
vehicles to improve claims adjustment
and specify body shop repairs.
 Automakers use AI to anticipate
needed maintenance and repairs,
and to suggest that customers go to
dealerships for that work.

doux, Mitchell's vice president for global
product strategy. "We strive for a repair performed according to OEM standards."
It's just a "question of time," Baudoux says,
before automakers and dealerships become
customers for similar programs.
Toyota Motor North America is working on
AI solutions in the cloud, a storage system that
encourages data sharing. Toyota uses predictive analytics to forecast trends in vehicle service, says spokesman Corey Proffitt.
The automaker reviews data to let customers know when their vehicles are due for service, whether routine maintenance or a recall,
Proffitt says.
At the same time, the system notifies a customer's dealership to ensure it has the parts in
stock to complete the repair.
General Motors' OnStar system uses AI to
notify customers when a starter, fuel pump or
other component may be going bad. It analyzes and monitors billions of pieces of data,
to determine whether a part is not performing
as expected, and automatically notifies customers if they need service, says GM spokesman Vijay Iyer.
OnStar also lets dealership service departments know of problems identified by the AI
system, Iyer says. 

Table of Contents for the Digital Edition of Fixed Ops Journal - February 2018