ChatGPT has been one of the most talked-about recent disruptors of the AI world. We have decided to put its knowledge to test and “invite” it as a panelist to our own virtual 'AI in Automotive Service' Roundtable. I interviewed ChatGPT along with Tilak Kasturi, Founder & CEO of Predii, and Mark Seng, Aftermarket Expert and VP of Business Development at Predii. All ChatGPT answers are exactly as shown by chat.openai.com - have fun reading!
Nadja: We’ll start with a basic question to ChatGPT - just to test the waters. ChatGPT: What is the automotive aftermarket?
Chat GPT: The automotive aftermarket refers to the market for replacement parts, equipment, and services for vehicles, including those that are no longer under warranty. It includes a wide range of products and services such as replacement parts, performance parts, accessories, and maintenance and repair services. The aftermarket is a large and diverse industry that serves both consumer and commercial markets.
Nadja: Not bad for a start. Mark, the Automotive Aftermarket has been your domain for 35 years - what do you think? Is there anything you would like to add?
Mark: Well, ChatGPT does provide a pretty good overview of what the Aftermarket industry includes - the overall parts and service business for any and all vehicles on the road or off. The Aftermarket, however, is pretty broad and deep in its entirety. What we refer to as the ‘Aftermarket Ecosystem’ includes everything that happens with that vehicle once it crosses the dealership sidewalk. This extends beyond general maintenance and repair to include parts manufacturing and distribution, collision, tires, glass, roadside assistance, insurance and much more.
The Aftermarket is then often separated into two broad channels: 1) the Independent Aftermarket - as represented by businesses like NAPA, AutoZone, Advance Auto, Christian Brothers, and Bridgestone/Firestone; and 2) Aftersales - which includes all the service and repair done by OEM dealerships whether or not the vehicle is under warranty.
And don’t forget that vehicle repair and service extends well beyond passenger cars and light trucks. The overall Aftermarket includes the maintenance of well over 300 million vehicles in the U.S. alone across medium & heavy duty trucks, buses, motorcycles, powersports - any type of vehicle really. They all need maintenance and repair services over their lifetime, and it all falls under the umbrella of the Aftermarket.
Nadja: Wow. This certainly sound more complex than the 3 sentences summary Chat GPT gave us. Building on this - our next question, again to ChatGPT: What are key challenges in the Automotive Aftermarket?
ChatGPT:
Increasing competition: With the number of players in the market increasing, it has become increasingly difficult for companies to differentiate themselves and stand out.
Changing consumer behavior: Consumers are becoming more price-sensitive and are increasingly looking for ways to save money on vehicle maintenance and repair.
Advancements in technology: With the rapid advancement of technology in the automotive industry, aftermarket companies must continuously invest in research and development to stay current and offer the latest products and services.
Online sales: The rise of e-commerce has led to more consumers buying automotive parts and accessories online, which has reduced the number of customers visiting traditional brick-and-mortar stores.
Supply chain disruptions: The global supply chain for automotive parts and accessories is complex and subject to disruptions due to factors such as natural disasters and trade tensions.
Environmental concerns: The automotive aftermarket industry is facing increasing pressure to reduce its environmental impact and promote sustainability.
Nadja: Sounds reasonable - and pretty complete. Mark, would you mind adding a little more depth once more?
Mark: This does nicely cover the main challenges facing the aftermarket business today at a macro level. There are some consistent themes across the 6 items including the acceleration of complexity while at the same time an increase in consumer “self-service” with access to more information than ever before.
But I believe the subject requires one to dig deeper. The challenges are pretty unique across the various segments of the industry. They are a bit different for manufacturers, distributors, OEM dealerships and repair providers. I would like to focus and add some detail on the challenges facing perhaps the most critical segment of the Aftermarket - maintenance and repair - where the consumer has the most interaction and exposure to the Aftermarket ecosystem - where the part gets installed and the box gets thrown away.
Nadja: Could you explain what you mean by that? Why is this part of the value chain so critical?
Mark: The majority of the challenges facing this final step of the supply chain impacts all other segments of the market. The decisions made by the repair technician - as well as the speed and accuracy of those decisions - impact the entire ecosystem.
One of the biggest challenges is the accelerating complexity of vehicle technology. Moving from traditional ICE powertrains to hybrid and EV vehicles has brought with it an increase in electronic sensors, computers and interconnected systems. What was once a simple replacement of a single part may require calibration of another seemingly unrelated component. This requires a new level of training and increasingly more complex diagnostics. All of this is happening at the same time as we are seeing a shortage of technicians across the industry. The technicians in place need all the help they can get with diagnosing the problem quickly and accurately in order to turn the bay and begin working on the next car.
This brings a unique challenge for the independent aftermarket specifically. As vehicle complexity increases, many consumers may feel their only option for repair is the OEM dealership. The independent aftermarket must educate the consumer that they do have the parts and experienced technicians required to diagnose and repair their vehicles correctly. This is very much like back in the 80’s when vehicles moved from carburetors to fuel injected systems. The independent aftermarket survived that evolution by educating the car owner - they must do it again.
Nadja: Thank you. As you mentioned initially, we tend to look at the Aftermarket as separated into the Independent Aftermarket - which you just referred to - and the OEM side of things, Automotive Aftersales. Which challenges do you see here?
Mark: The OEM dealerships have their own unique issues as well. They are not always seen as the cheapest alternative for vehicle repair. As repair costs for the more complex technology goes up, the consumer is also gaining more access to data and information. The interaction between car owners and their dealership repair shops will be forced to become more transparent as it has in the independent aftermarket.
The impact extends beyond just the diagnosis and repair of the vehicle. The complexity has added many more SKU’s that need to be managed and stocked which is stressing an already difficult to manage supply chain. In addition, those parts are, on average, more expensive as well driving an added urgency around getting the right parts to the right place at the right time.
All is not lost, however. The Aftermarket has one great advantage in facing these challenges. The industry sits on a growing treasure chest of data. It owns repair order and sales transaction history of all the make, model, engine combinations being repaired. There is more and more data being provided by diagnostics, calibration and alignment systems, vehicle sensors and connected car data that can all be used to accurately and quickly diagnose the problem as well as predict what repairs will happen at what mileage during the lifetime of a given vehicle.
If the industry can be successful in making sense out of this large amount of data - standardize and make the correlations needed to leverage its value - the Aftermarket has the potential to meet and overcome these challenges.
Nadja: Thank you, Mark. You brought up data as a huge opportunity for the Aftermarket Ecosystem. Since this is obviously a topic of interest and expertise at Predii, I’d like to discuss how Data and AI come into play and how AI specifically has the potential to disrupt the Automotive Aftermarket. Let’s start at the beginning. ChatGPT, can you start by defining AI for us? What is Artificial Intelligence?
Chat GPT: Artificial Intelligence (AI) is a branch of computer science that deals with the creation of intelligent machines that can perform tasks that typically require human intelligence such as visual perception, speech recognition, decision-making, and language translation. The goal of AI is to develop algorithms and computer programs that can process vast amounts of data and perform complex tasks with a high degree of accuracy and efficiency.
Nadja: Tilak, you’ve been in the AI Business for years. Is there anything you would like to add? AI is obviously a complex subject - would you mind sharing what you understand by ‘Artificial Intelligence’?
Tilak: I will try not to give a computer science definition which you can get anywhere - and ChatGPT just provided that for us - however, I will try to answer in the context of a business organization or enterprise. AI in an enterprise specifically is the ability to use data to learn to emulate decision making of a person who in the absence of AI has to use his/her experience and available “data” to make the same decision. In other words: AI in a business context augments human capabilities.
Nadja: Let us stay here for a moment, specifically on the decision-making. Chat GPT, how do you think AI can be applied in the automotive aftermarket?
Chat GPT: AI can be applied in the automotive aftermarket in several ways, including:
Predictive Maintenance: AI algorithms can analyze data from sensors on vehicles to predict when maintenance is required, reducing the likelihood of breakdowns and improving fleet uptime.
Diagnostics: AI can be used to diagnose issues with vehicles by analyzing sensor data, error codes, and other information, reducing the time and cost of repairs.
Inventory Management: AI algorithms can help automate inventory management by analyzing sales trends, predicting future demand, and reordering parts as needed, reducing waste and maximizing efficiency.
Customer Service: AI-powered virtual assistants can provide quick, accurate responses to customer inquiries, improving the overall customer experience.
Price Optimization: AI algorithms can analyze market trends, competitor pricing, and other factors to optimize pricing for aftermarket products and services, improving profitability.
These are just a few examples of how AI can be applied in the automotive aftermarket, and as technology continues to evolve, it's likely that new applications will emerge.
Tilak: Let me add to that. Some of these terms like Predictive Maintenance, AI powered virtual assistants, etc. are not new in the industry. What is state of the art here is the ability to apply AI in a cost-effective & service operations friendly manner and not like a science experiment.
To do that, there is one more, very essential step required, which is data normalization. As Mark explained earlier, the Automotive Aftermarket is extremely complex. The data we are able to base these solutions on is data is rich, but it’s also messy, noisy, and unstructured. Data Normalization. It is the foundation that allows us to build extended models for predictive maintenance, diagnostics, etc.
Nadja: Understood. In order to develop, let’s say a Predictive Maintenance solution, we first need to do some housekeeping and clean up the data foundation. Could you elaborate on how AI can be leveraged for that?
Tilak: Absolutely. Let me use an example: If I were a product manager for brake pads, I would like to know why, when, and what quantity of my brake pads are being sold in the market. Parts ordering is one source - but it’s incomplete. To really understand how many of my brake pads ended up on a vehicle I would need to look at repair orders. Now, the tricky part about a repair order, if you’ve ever seen one, is that no RO looks like the other. What one mechanic might call a disc brake pad would show up as a ceramic pad, or a disc pad in the next repair order. AI can support here by acting as a ‘scalable expert technician’ who can scan through millions of repair orders, understand that disc brake pad is the same as ceramic pad and normalize all variations to one searchable, actionable term. By continuously learning from large amounts of automotive service data, AI can build a foundational knowledge base to augment decision making, and support - for instance - market insight for our brake pad product manager.
Nadja: Thank you. I would like to get back to Mark’s statement from before: Maintenance & Repair is perhaps the most critical step in the whole process. How does AI help here?
Tilak: AI can be leveraged across the entire value chain because of its ability to understand correlations and patterns across large amounts of data. Once we’ve called a ‘brake pad’ a ‘brake pad’ or we know that ‘P0300’ is an ‘engine misfire’ we can begin to analyze and use correlations across different data sets.
Imagine a driver - we’ll call him John - is driving his 2019 Honda Odyssey Touring mini-van. John is unhappy because he keeps hearing a squeaking sound from the front left of his van when he is taking a turn or braking. He calls the shop and takes his vehicle for a check-up where a service advisor writes up a work-order and, upon inspection by the technician, recommends replacing front and back brake pads. John approves the repair order and gets the service done.
An AI system that is capable of understanding the automotive domain will learn from all data touchpoints in this ecosystem - repair orders, ERP systems, parts ordering, even the initial customer complaint. We can then feed this knowledge back to the technician in form of recommendations - vehicle specific, based on millions of repairs. Further, imagine an AI system securely monitoring (with permission from the end-user) the critical components of a vehicle for safety: like airbags, brake module, battery, coolant and allowing the shop to pro-actively call the customer…
Nadja: …so John never needs to be unhappy to begin with. Thank you. I think this was a great conversation - I would like to do a follow up on this and dig a little deeper into the ‘how’ of AI solutions. Stay tuned! Mark, can I ask to for a closing statement? Is there anything else you’d like to add?
Mark: Well, AI is certainly well positioned to assist the Aftermarket with the many challenges it faces - and there are many. However, it all comes down to the one thing the industry has struggled with for as long as I’ve been around - getting the right part to the right place at the right time. Maximizing both supply chain efficiency and customer satisfaction in the process. Based on how Tilak describes the application of AI within the Aftermarket ecosystem, we have the opportunity to go beyond that goal and insure the right parts are in the right place - ahead of time!