The $50,000 Breakdown That Never Happened: AI for Truck Shops

**[00:00:00]** Hello everyone, my name is Amando Presceno and welcome to the web talk show. Today we're talking about AI for truck repair shops, truck maintenance shops. What if I told you that you could predict a catastrophic failure three months before it happened? So that's the topic of today. If you have any questions, please load them up in the chat either on LinkedIn where we're streaming or on Tik Tok where we're streaming as well. Now, this topic of conversation for today specifically started because I did a few auto repair shop related events recently and someone asked if this would apply for a truck maintenance shop as well. It definitely does and I'm going to give a few examples of how AI can be used in a truck shop, truck repair shop, truck maintenance shop, anything related to trucks in general. There are so many ways that this could be applied and it's so much more powerful even than the just car repair industry because if you think about it in the auto repair industry while it's extremely important for us as consumers, regular daily drivers and the care of our families expenses in these and costs of the car not working, things like that are going to be hate to say but minimal compared to the costs of trucks failing. Why? Because trucks, apart from potentially being an accident factor, which would be humongous, have halls. [laughter] So, so trucks can be carrying thousands, hundreds of thousands, or even millions of dollars in their rear end. So, it's very valuable to keep them sharp as attack. So, that's what we're going to be talking about today. And I'm going to give you a few scenarios or examples that apply in this particular industry and how you as **[00:02:00]** a shop maintenance owner or even as a truck driver wanting to have something to be better prepared can consider. Let's start with just a basic example of something that happens often, which is that phantom vibration, right? You're on the road. You you're you're driving. You feel something. Some something feels off. You're not quite sure what it is and you go to the shop and the just I don't know, maybe you went for an oil change or something or you decided to take it just for that because you don't you just there's something you don't feel comfortable with it. But then the texts analyze it. They put the machine in it and everything and they're like, "Nah, I don't think it's anything. You shouldn't be worried about it." Now, if you had AI in the shop looking at not only the source data from the truck readings, but also at vibration patterns, cross referencing with previous engine failures from other trucks that are the same brands, same model, or even historical things that you've done in that particular shop. that AI can then look at that with the additional information and come up with a conclusion that this might mean that XYC is happening, right? So maybe it'll have an 87% probability that the turbo actuator is going to fail within 800 miles. So if you catch that with AI because AI can keep all the data on hand and make that assessment quickly apart from whatever your techs are doing and that would mean what $400 that you spend replacing a part than a $12,000 bill for the roadside to emergency repair and then of course the missed delivery penalties. So, [laughter] there's a lot you can do here to just have a better handling **[00:04:00]** of potential failures. I talked in the previous episode about how even a proactive approach to things, even if your customer didn't bring the car in, speaking of the car repair shops, your system could be looking at information about recalls or issues, common issues with a specific brand or model or year of a particular make. and proactively send out messages to your customers saying, "Hey, we know the Ford Bronco has this issue within 30,000 miles. You might want to bring it in. We could just look at it. Everything's fine. You're okay, but maybe we want to replace this little thing." And this is very useful when you come into especially cars that are outside the their just manufacturer warranty period. So bringing it towards the truck side or the truck maintenance. Can you imagine these truck maintenance shops sometimes have fleets that have span hundreds thousands of vehicles and so they have so much data even in house not even talking about thirdparty providers but in house they have so much data of everything that's failed in the past what type of trucks what type of beds what type of loads uh the periodicity of when they change the oil type of fuel they're using what additives they're saying which ones failed, which ones had this issue or that issue. So, if you can imagine, there's a system that has AI that's just going through the data for the trucks that need maintenance that are coming up. Maybe you already have everything scheduled. People are coming in. So, instead of just checking all the trucks, you can just see, okay, so these are coming into the shop anyway, let me review those first. And then you can start to spot trends and say, "Oh, it seems **[00:06:00]** like this is failing for 90% of the trucks for this combination of engine and configuration. Well, maybe we need to look at it." Then they could abstract and see other information. I could research additional data and find that in a certain model build. Turns out the connectors weren't correctly soldered or they used the wrong gauge wire. This stuff happens in real life, not only in trucking and vehicles. It happens [laughter] in general. Um we've had when when I was in the theater space sound engineering uh and and studio work there was this console Mackie console that works great but that all of them had an issue with the goldplated contacts inside. So there was there was a board that was connected. this is this is an electronic digital board and all of them had that issue. But most people just thought after a while the console just died and that's it [laughter] and they just sold it on eBay or whatever. But a friend of mine who's genius with all this stuff, he figured it out and so he fixed his own and then fixed all the other ones that he could find and amassed a great library of these machines that work fantastic as long as you know that that issue happens. So you can imagine this happens with vehicles as well. So if you already know what something is going to look like when it fails and what that failure looks like, you can save a lot of money figuring out beforehand what that's going to be and figuring out which of your fleet need that adjustment earlier and you can save a lot of money. Another thing is we have my notes here and I wrote the part psychic but it's let's **[00:08:00]** say it's January you're in Chicago and AI can look at your historical data like we were talking about right now the weather forecasts it could look at the fleet age distribution and it could tell you order 40% more batteries and why well because you know historically in that time period, you're going to need a lot more batteries because they're going to break down because of the cool because of cold or any other thing, right? So, maybe you need more batteries. Maybe you need more block heaters. And it can even tell you when you need them. Just order them. You need them by 3 weeks. So, you already have everything ready when your competitor is going to be scrambling for parts. And this is something that Walmart and other huge commerce businesses have done for many years. There's a lot of data that can be used to your advantage. And before maybe you needed to be the size of Walmart to do something like this. But now you can with the relatively small investment understand and analyze all your data with AI. And if you're a trucking company or a fleet or you have a ton of maintenance shops, you can use that data to your advantage and make sure that your whole fleet is running the most efficient way it can. So figuring out which parts you need beforehand, figuring out which trucks might have an incident before the incident occurs, figuring out what the diagnostic for something is even before upscaling it to your high level texts because I imagine you have your entry-level text and then you have your senior texts and a lot of time your senior texts are busy with the most important challenges. ing situations within the shop and you **[00:10:00]** don't want to bother them for every little thing. Now imagine that you have their knowledge because obviously they well some of it is stored in your systems already. Some of it you can have them sort of jot down and create your own knowledge base but a lot of it can be abstracted from the things that you've done in the past and notes and all your um order sheets and everything. So that when a junior tech is working on something and they sort of can't figure it out yet, they could have an agent using a phone or a little microphone like one I have here just tap it and say something like, "Well, I'm working on this Cummins TSB 2024, whatever." And then the issue seems to be that the spark plugs are behaving this way. what could it be? And then the AI could then interpret that, look at the knowledge base of all your data and then bring back a response in plain English in whatever grade level you wanted to and give it to the person directly. You could even respond, maybe they have their hands dirty, they're full of oil and grease and they could just hear it directly in their headphones or out loud. You could tell it, "Oh, make sure you check the levels on this particular thing or this other item." and then they could just check that off and maybe they find the issue very fast without having to go and bother one of the senior techs. So that's another little thing that you could do there. If you look at other examples, we could see warranties, right? So let's say the tech finishes a repair, they snap a photo of the failed part and then the AI **[00:12:00]** can scan it. We've done things like these before. I could cross reference with the OEM bulletins say this is a failure that matches this particular pattern. There is a warranty claim eligible and I've already submitted the draft. There's then you could just go and check it make sure that it's okay or it could tell you that it's there's a recall already in place that you can make sure that you cover. There's also items that you can handle like the base. So I imagine you might have a lot of base and a lot of the trucks might need service at the same time. So the AI could look at the job complexity and parts availability and then it could go in and find which texts have which certifications and auto assign the text to the base to the jobs based on their historical times, how fast they are and [gasps] what they can do, what they're what they're best at. And that way your best diesel guy isn't on the simple job. he's just done the best jobs. And a lot of you might already have something like this in place, but if you don't, then maybe it's too expensive to do that now with AI. It just democratizes these tools that you can use to then empower your team to work in a more efficient way. That's what I like about AI. It's a tool that we can use to make things that were very complex before a lot simpler. Sure, you can make mistakes and AI can make mistakes, but there are ways to make it work properly. If you work with a professional, you can work with us, you can work with any other AI company. And what they'll do and what we'll **[00:14:00]** do is first do an exploration of what you have already, which systems you use and then from there we can make a plan of what things could be automated, what data can be pulled for the knowledge base, which are the third-party systems that we should interact with. How what is the way that your drivers interact with the shop? Do they use WhatsApp? Do they use voice? Do they call? Do they text? Do they use the radio? How can we interface with them? Because at the end of the day, what you want to do is make it simpler for everyone, for the manager, for the techs, for the drivers, for the owners. You want to have all the data available for each one where they are. You don't want to give them a new interface, a new app that they have to figure out how to use and where to tap and where to click. You want something that they're used to. If they're used to using voice on their headsets, then give them something that works with voice. Voice AI is super advanced nowadays. you can just talk to it and it'll talk to your different systems and it can pull the data and it could give them data back. It could help you with scheduling and a bunch of other stuff. If they're used to using messaging via text, then meet them where they are, create an interface that allows them to just communicate with the shop via text or via WhatsApp or via Telegram. It really varies depending on the text and on the drivers and on the CEOs. Where are they? And that's where you should meet them when you're planning a system like this. And remember, a lot of these drivers **[00:16:00]** are working cross borders. So, does it make sense for them to be using regular landlines or maybe void lines or does it make more sense for them to use a tool like WhatsApp or Telegram? What are they using now? And then you can interface it into it. Now, one of the great things is that most of what I'm talking about right now can be multi-platform. So you can have that knowledge base and then just expose it in different platforms. Expose it as a text agent or a voice agent for the CEO to get certain dashboards about what's going on, how the shop is running. Or you can expose it as a voice microphone based voice app for your text when they're working on the different trucks. Or you could have it where it's an email-based interaction where someone sends an email and then that email gets processed put through the system and then a response get generated and the email gets sent back. So that works for parts for recalls for invoices for processing all the different things that I've talked about can be interacting with each other but then also you can access them in different ways. And that's the great thing about it. It's no longer tied to, oh, I can only use it on Windows. I can only use it on Mac. I can only use it on an iPhone or on the web or on this archaic device that the manufacturer gave me. No, you can have everything tied together and then use it in all sorts of different platforms, voice, messaging, text, web, whatever you want. Streaming, video, streaming, audio. Can you imagine if you had something on the truck that's just hands-free and your drivers could just interact with **[00:18:00]** it and ask it things about how the truck is running because I imagine you have a bunch of sensors on it. So could ask about things about maintenance, about when it has to stop, about um cooling, heating, how everything's going, if the load is properly [snorts] set up, if they need to make an adjustment because they're getting a new load. A lot of things can be done via voice now. And that's where it gets really, really interesting. Another thing that we can talk about in this space is the service. So, let's say the fleet manager calls in at 3:00 a.m. a truck broke down in El Paso. So, instead of voicemail, if you have, speaking of the voice AI agent, the voice AI agent can pick up, pull the truck maintenance history, walk the driver, not another tech, just the driver through basic diagnostics. it because dispatch mobile service and then can update your system all automatically before most of the rest of the crowd wakes up. Again, that's great because sometimes it won't even need the dispatched mobile aid. Maybe it's something that the truck driver can figure out right there on the road because it's something that's common and it's something that can be easily fixed. Maybe they just need a specific part that they could get at the next truck stop. Or maybe it is more complex and then it should automatically dispatch someone to meet them at the next truck stop. Or if they cannot move, then it'll automatically dispatch it beforehand. So even if the truck driver is trying to adjust it or fix it and maybe gets it done, someone is there to meet them as soon as possible. In this case, the soonest possible because nobody had to **[00:20:00]** wait for a dispatcher. This could work 24/7 on autopilot. So, there's a lot to do here. There's it's a space where compared to traditional autoshop, there are numbers. So you it's more likely that there are fleets. There's going to be a lot more data and the outcomes of if something fails are sometimes greater both in monetary sense and also in health and safety because we have to make sure that all the trucks are running safely because they're huge and heavy and they're driving on the same roads as regular passenger cars. We want to make sure that everything is working fine. And this is why it's so important to make sure that everything is in tip-top shape. So, I'm going to go through a few other examples. And if you're on the live stream, because this is live streamed apart from recorded, you can comment live on LinkedIn. If you're watching it on LinkedIn or on Tik Tok as well, you can also comment live here and I'll take any of your questions at any time if you have a question. Otherwise, if you're watching it on YouTube or on Spotify later on after the recording, you'll you're free to also leave any comments there and I'll be happy to answer them on another live stream or directly on the comment itself if the platform allows for it, of course. So, I'm going to talk about a few other examples. Let's talk about the translator. So the driver sometimes is talking in their language, whatever they can express themselves with. So the driver is going to have a different education level than the tech. Maybe they have the same education level, maybe they're used to same things, maybe they're used to different things. And so the **[00:22:00]** AI can also translate between talk. So this is something I'm very good at specifically when I'm talking with business owners and tech tech related providers and sort of doing the uh in between communication and translation. Translating technical jargon to a regular person or to business speak is sometimes difficult and it's really helpful to make someone or help someone understand something that's very technically complex. The same thing happens here. So, the driver might just make something like, "It's making a weird noise when I go uphill and it smells kind of funny sometimes." So, the AI can take this vague mess of notes and cross reference it with the vehicle's age, mileage, and common complaints, and it can output something likely DPF regeneration issue, check exhaust back pressure sensor, and doc efficiency. Estimated diagnostic time 45 minutes. And then that can go to the shop. >> [laughter] >> So now the shop knows what it's more likely that this person is trying to refer to. So that's where I'm going with this. It's the whole idea of using these tools not as a replacement but rather as an additive tool that will enhance the performance of your team of your drivers etc. Another one is going to be maybe regarding fuel. So maybe their fuel economy just suddenly tanked. So the AI can pull Telmatics data. It can compare it against similar trucks in the system and finds, hey, unit 47 is running 18% below the estimated or average miles per gallon. There's data showing that the injector number three is showing irregular pulse widths, which correlates with a pattern seen in other 12 Detroits before the injector fails. So you can catch it before that happens and save another $4,000. The examples are endless. You just **[00:24:00]** can do so many things here because you have the data. Let's say someone comes in and they call in with a quote on a clo quote uh for a clutch replacement, right? So AI can pull labor times from your actual historical data and then it can factor that in gets the specific trucks known quirks checks parts availability for you and then across your different vendors gets all the data it needs generate an actual accurate quote in 30 seconds and there's no more of that um let me call you back and then that person forgets and never calls him back and [laughter] it goes back to why sales are lost in the first place. So, there's so much again you could do with this um type of tooling that will make your shop run more smoothly. So, if you have a truck maintenance shop, if you work with trucks, if you have an auto repair shop, you work vehicles, there's so much you could do. And if you're wondering how this could make your shop more efficient, please reach out to us. You can send us a message. You can visit us on piscarreno.com. You can send us a message on LinkedIn, on X, on Instagram, on Facebook, on Tik Tok. We're available everywhere. Send us a message and I'll be very happy to walk through you with you through your company and see where are these special scenarios where you can have a quick win and see, okay, perfect. This works for me. Now, what else we can do is there's so much so much you could do to save time, save money, and make your shop work more efficiently. So that's it for today's show and if you like it and you haven't subscribed, **[00:26:00]** please subscribe to the podcast on Spotify or on YouTube or on iTunes or wherever you are listening to this podcast. And if you're watching live, thank you for watching and I will