2023-01-26 AI Committee Meeting Minutes
Â
Public Page
Â
Â
Â
Date
Jan 26, 2023
ANTITRUST STATEMENT
As participants in this meeting, we need to be mindful of the constraints of antitrust laws. There shall be no discussions of agreements or concerted actions that may restrain competition. This prohibition includes the exchange of information concerning individual prices, rates, coverages, market practices, claims settlement practices, or any other competitive aspect of an individual company’s operation. Each participant is obligated to speak up immediately for the purpose of preventing any discussion falling outside these bounds.
Agenda
Welcome/Networking
Antitrust
This meeting is subject to the terms of our anti-trust statement shown here. In addition, this meeting may be recorded. Is everyone in agreement
Meeting minute Review
Review Raj Document
Look at BMS Data
Meeting Minutes
Antitrust Accepted
Meeting Minutes Reviewed Accepted
Meeting minutes:
We need to look at how we accommodate different industry segments with AI technology. We are looking at FNOL, but we’re taking a myopic view that is going to be FNOL for an appraisal.
We can see AI creating very detailed and accurate paint estimates and be able to facilitate that into ordering the paint.
How will recyclers use the data from AI?
FNOL is a legacy term we have used for decades, if not a century. But FNOL is changing with the new technology and the actors are changing. The car may be making the initial FNOL. We need to make sure we are capturing all the data and mostly likely have a class that says ‘WHO’ is making the FNOL request (Customer, vehicle, insurer, etc.).
FNOL does have a code list we can use to say who is making the FNOL request, as well as AdminInfo that has the different actors for the message.
Reminder we are looking at the BMS for FNOL to look at the types of data captured today to help us look for data we may need that is missing.
We do want to walk through a complete lifecycle of an accident to see where AI will impact the industry, the FNOL, estimate, repair, salvage, the FNOL was just the first one the committee was looking at.
The AI committee was created by CIECA because we knew AI would impact a number of the CIECA Products, if it was only going to impact FNOL we would have activated the FNOL committee and they could have handled the changes, if it was only going to impact Estimate, the Estimate committee would have been activated. Due to AI impacting many products, the AI Project Committee was created. This Committee will look into the impacts and recommend changes to the Products and messages. The recommendations will be reviewed by the Product Committee for experts in that message to review, then the SAB/Architecture Committee’s will review any changes.
FNOL is a catch phrase in the industry. With the FNOL possibly coming from multiple venues, we may be able to change the terminology to VNOL for Vehicle Notification.
The vehicle sensor or camera could be the first to determine that an accident has occurred.
Could be a watch or and external component of the vehicle that makes the first notification.
FNOL was term that goes back to the paper form, those generated by a group called the Accord.
CIECA could drop the FNOL call it NOL?
There are many types of Notification Loss message.
Traditional ICE vehicle vs electric, you know battery operated electric vehicles is a different source.
It would be good to know that the sensor of a vehicle thought it was an accident, when in fact it may have been a snowplow hitting a curb and no accident at all. So good to know the source.
What sources, the sensor data versus a traditional CMS data ingestion endpoint. All AI partners have to be able to consume that as a training function for the AI because it can inadvertently become a little bit biased.
Challenges to try to avoid and best practices for the standards body is what we can look into as a committee.
We do not want our Best Practices to restrict the user in the type of format or how many pictures that can be sent.
metadata is important to determine the best practices. The minimum can always be sent, it will just impact the AI function. For the best results, the best practices should be meet.
AI Input vs Output
The AI systems will require pictures, videos, VIN, Make, Model, Year, etc. to be able to run through the equations and come up with an estimate of damage.
The output of the AI systems that goes to the industry will be different, what does the recycler need, what does repair shop need?
CIECA needs to look at the standard payload for a request and a standard payload for a response.
The payload for Acquistion would be a good place to start.
We have a number of Use Cases that will be identified, however, it may be possible that the payload will be the same for AI.
Up Next
Welcome/Networking
Antitrust
Meeting minute Review
Jimmy is going to show a demo of how AI works.
Review Abhijeet document to go over the payload for request and response
Review Raj document
Action items
Decisions
- Committee agreed that the data needed to be able to identify who was sending it, we need to look at new areas of technology, such as sensor data being added to the code list or a new data field.
Â
Participants
@Paulette Reed
Ginny Whelan
Stacey Phillips
Jeffery Hodge
Paul Barry
Jonathan Pyle
Abhijeet Gulati
Don Porter
Ken Eagleson
Olivier Baudoux
Kimberly Sorlien
James Spears
Bryan Williams
Timothy Ronak
David Willett
Participants in the meetings are noted for your information. If you have questions on the committee’s activities, please contact a recent attendee. AI Committee Home
Â