00:00 00:00


Are Fleet Safety and Productivity Technologies Worth It in 2023?

Learn about the technologies fleets and heavy-duty companies are using to increase safety and productivity.

Episode 258: In this episode, the subject is all about safety and productivity. We have two industry experts to discuss this subject from two very different perspectives. First off, we will learn how Boyle Transportation approach how they took to make their fleet safer for their drivers on public roads, and the steps they took to achieve a ZERO accident rating.

Additionally, we also got the chance to talk with SpotAI about how they use AI technology with video to help businesses to become safer and more productive.

Michael Lasko the Director EHS and Quality at Boyle Transportation

Rish is Co-Founder and Head of Product at Spot AI

Learn about the technologies fleets and heavy-duty companies are using to increase safety and productivity.

Watch the Video Version

Disclaimer: This content and description may contain affiliate links, which means that if you click on one of the product links, The Heavy-Duty Parts Report may receive a small commission.

Sponsors of this Episode:

  1. Want to look up parts but don’t have a part number or the VIN? Download Diesel Parts for free on Desktop or on your Apple or Android device.
  2. Ambac International is hiring! Apply Now
  3. Looking for suspension, steering, and 5th wheel parts? Buy now from SAMPA.com
  4. Looking for OEM genuine parts? Buy now from AlliantPower.com

Buy Parts:

Are you looking to purchase heavy-duty parts and get your commercial vehicle repaired? Get access to one of HDA Truck Pride’s 750 locations across the United States and Canada. Visit HeavyDutyPartsReport.com/BuyParts.

Transcript of Episode:

Jamie Irvine:

You are listening to The Heavy-Duty Parts Report. I’m your host, Jamie Irvine, and this is the show where you get expert advice about heavy-duty parts that keep trucks and trailers on the road longer while lowering costs per mile.

In this episode, the subject is all about safety and productivity. Welcome to The Heavy-Duty Parts Report. We’ve got two people that we’re going to talk to today. They’re both industry experts in their own rights and they’re going to talk to us about safety and productivity from two very different perspectives. So our first guest is Michael Lasko.

Now he is with Boyle Transportation. And what we’re going to learn from Michael is we’re going to learn about the approach they specifically took to make their fleet more safe for their drivers on public roads. And we’re also going to learn about what steps they took to achieve a zero accident rating. It’s very interesting the approach, and I think there’s a lot to learn from Michael and what has been done over at Boyle Transportation.

Then we’re going to talk to Rish Gupta from SpotAI. Now he’s got a very different perspective because he’s looking at the trucking industry from a technological perspective, from an AI perspective, and applying that AI technology to video and using that to help businesses in the transportation industry in trucking to become safer and more productive.

So his perspective is one that I think you’re going to appreciate and it’s really exciting to learn about some of these technological advancements with artificial intelligence and how it’s being deployed inside of the trucking industry. But first, let’s talk to Michael Lasko from Boyle Transportation.

Michael Lasko:

Hi, my name is Michael Lasko. I’m the Director of EHS and Quality for Boyle Transportation.

Jamie Irvine:

Michael has worked in the trucking industry for over 20 years. He’s been in his current role for just over seven years now. Michael, welcome to The Heavy-Duty Parts Report. So glad to have you here.

Michael Lasko:

Great, thank you very much. I’m glad to be here.

Jamie Irvine:

So I’d like to know from your perspective, as someone who works at a fleet, how do you approach safety? Let’s just talk kind of an overarching view of that and then we’ll get into some details.

Michael Lasko:

So see, the way we look at safety and boil transportation is we really approach it from a risk reduction strategy. I know that there’s a very popular mission to zero in those kind of catchy terms that are in the industry, but unfortunately those don’t really get a lot of buy-in from your professional drivers. So really what we look for are opportunities where we can assess risk and then try to mitigate that risk to help keep our fleet safe, the motoring public safe while they’re around our trucks, and reduce the number of incidents that we encounter as a company.

Jamie Irvine:

So when you did this analysis, you looked at the number of incidents where you had an accident and you were actually able to get that down to zero by focusing on ADAS. How did that play a role? How did you come to that conclusion that was an area you needed to focus on?

Michael Lasko:

Yeah, so Boyle was really an early adopter of these technologies. As they became available, we started out with Lane Departure warning system and then kind of progressed into the collision mitigation with forward-looking radar with emergency active braking. So really what we found was is a little bit more involved with these systems than just putting it in your truck and telling your drivers, Hey, go figure it out.

And I kind of liken it to the process of when you buy a new car, you buy a new car. I don’t know too many people that have ever actually pulled out the owner’s manual and read all the details. And a lot of times, even if you do do that, you get just marketing messages, automatic emergency braking, it helps avoid collisions. Well that’s great, but it doesn’t give us any detail about how the system functions and what the parameters are.

So we identified that as a very early on as a training opportunity and we’re able to work with our professional drivers and show them the ins and outs and this is how the system activates and here’s the protocol for this and here’s what you can expect in this type of situation and false positives and all the things that come up with these systems. We prepare our drivers for it before they hit the road. So it’s not a surprise when they encounter a false positive.

This is usually something that drivers experience and instantly, I hate the system. I don’t want to want automatic emergency braking. It robs me of control of the truck. And really those kind of things are just drivers voicing their frustration with the lack of information that’s has been historically available with these systems. So long-winded way to answer your question is really just education. We train our drivers, we teach ’em how to interact with these systems, we teach ’em what they do and what to expect and that helps you achieve the results that you want.

Jamie Irvine:

That makes a lot of sense to me. And that proactive approach obviously worked with drivers. How did you approach bringing all of your equipment up to a certain standard? Because obviously you’ve got a fleet of various ages, and so that technology was coming in different phases. So did you do anything specific to bring your fleet up to a minimum standard?

Michael Lasko:

Yeah, that’s a really, really good question because you draw attention to an issue that I think a lot of fleets don’t fully appreciate, which is when we buy our trucks, we buy 10, 20 trucks at a time and we buy, fortunately these days, most of your manufacturers provide a kind of suite that you can buy where initially you had to piece meal these things together, so would you end up with 10 trucks of this variety and maybe the next year it was a very similar platform, but every once in a while there’d be a pretty substantial upgrade.

So what you end up with is somewhat of a mixed fleet that has the same kind of technologies that operate quite a bit differently. So we provide individualized training to our drivers, hey, this is the truck that you’re getting into, this is the level of this is how it’s going to respond in these situations. This is the level of activation when this happens. We make sure our drivers know that specific to the equipment that they’re going to be operating.

Jamie Irvine:

That makes a lot of sense. When it comes to getting support from the manufacturers of this different kinds of equipment, how did you approach that? Was there areas where you felt the support was really strong and other areas where you wished that there was something else? What was the journey like through that? Because I can’t imagine you were able to create this proactive system all on your own.

Michael Lasko:

It’s really interesting. So back in 2015, 2016, like I said, we had kind of a hodgepodge of different equipment manufacturers, whether it was roll stability or lane departure, collision mitigation. I actually think we had three different brands of all of that. To answer your question, it was very difficult in the very beginning, driver’s voice concerns mostly about false activations or false positives, and it was really difficult to get drivers to buy in to the system. At first, a lot of pushback.

The game changer was video. We went with a video event recorder inside of our trucks, and we were able to start capturing a lot of these false positives that the drivers were complaining about. And initially when we’d reach out to the manufacturer for whichever one of these systems that was kind of acting up at the moment, we would generally just get a canned response about, Hey, your drivers just don’t like this.

They don’t like that the truck’s going to apply the brakes and it takes some getting used to. And then so I said, well, why don’t you guys come in and have a meeting? So then I showed them, I had about 600 videos queued up, just, Hey, let’s watch some videos. And we sat there and made these guys watch. I want to say we watched 10 or 15 videos.

Jamie Irvine:

I was going to say 599 of them.

Michael Lasko:

We didn’t have to go that far, but we watched 10 or 15 videos before the guy that was there that the sales rep was, and actually I don’t think it was a sales rep, I think it was an engineering guy, said, okay, clearly there’s a problem here. And I don’t know if it’s true or not, but I think we’re probably one of the first that really stumbled upon the need.

When we talk about maintenance of these systems, firmware updates and software updates, those are all things that have to occur. Instantly we started getting firmware updates from the manufacturer and it reduced the number of false positives, and we had a lot of success with that. And then as the years progressed, we kind of went with a package suite bundle from the manufacturers that we purchased our trucks from. So in the very beginning, it was very challenging.

It was tough to get any kind of actionable response out of the manufacturers. And I think over time they’ve kind of learned those lessons and they’re a little bit more forthcoming with, Hey, this is what you have to do to maintain it, and you’re going to have to train your maintenance folks, and you’re going to have to make sure the drivers know how to indicate when something needs to be repaired and what the problems are that they’re experiencing so that maintenance staff can then say, hey, maybe this is the firmware update or maybe it’s a calibration issue.

One of the big ones I always like to point out is a lot of these systems they talk about is self calibrating. And that’s true right up until the point where it needs to be manually calibrated. And that’s the thing that we try to get our drivers to understand is sometimes these things have to be brought into a shop and manually calibrated so that they’re operating in optimal performance.

Jamie Irvine:

Yeah. Well, that’s fantastic. Thank you, Michael for taking some time talking with us and sharing the journey at Boyle Transportation. Congratulations on the achievement of zero accidents for the, what’s the time period that that’s been achieved in?

Michael Lasko:

Well, it depends on how you want to look at the data we’ve had in over eight years, I think we’ve had one preventable recordable accident. So we’ve been, unfortunately we share the road with a lot of folks. So we’ve been in other accidents that fortunately, non-pro preventable and very minor incidents. Nobody was hurt and all that good stuff. But yeah, we’ve had outstanding accident history with ADAS and training and preventative maintenance. All these things combined have really created a really good recipe for an amazing sauce.

Commercial break:

We’re going to take a quick break to hear from our sponsors. We’ll be right back. Don’t have a heavy-duty part number and need to look up a part? Go to parts.diesellaptops.com or download the app on Apple or Android to create your free account.

Looking for high-quality fuel injection for heavy-duty applications? Having one supplier for fuel injection allows you to better serve customers by providing them with a complete line which increases your sales and profitability. Learn more at ambacinternational.com/aftermarket.

Parts availability and quality have a big influence on fleets and owner operators’ total cost of operation, if they can’t find a part, it means more downtime. If they install a low quality part and it fails, it means even more costs like tow bills, hotels, meals for the driver and lost revenue. That’s why we recommend Sampa. They manufacture a wide range of advanced parts for commercial vehicles. Their website has an intelligent product search engine and broad coverage of suspension, steering and fifth wheel components. Expect More. Expect Sampa. Visit sampa.com today.

Jamie Irvine:

So we’re looking at technologies that are going to help trucking fleets to be more safe, and I wanted to bring on someone who has some innovative technology around the ‘videofication’ of everything that I think you’re going to find particularly interesting. So I’m happy to introduce Rish Gupta. He’s the Co-Founder and Head of Products at Spot AI. Rish, welcome to the podcast. Glad to have you here.

Rish Gupta:

Happy to be here. Prior to SpotAI, I was at a company called Samsara, which deals very closely with different trucking technologies, so very familiar with the space. And then with SpotAI, we are doing a lot video technologies across industries including the trucking space. So super excited to chat about it.

Jamie Irvine:

I want to learn a little bit more. So let’s start with what technology is SpotAI bringing to the trucking industry? How is it unique? What does it do? We’ll start there and then we’ll talk about some of the ways that it’s used to improve safety for trucking fleets.

Rish Gupta:

Yeah, so at a very simple level is think about all the security cameras or cameras which are there in loading peoples in your warehouses, in anywhere where the trucks are basically docking to load, unload, and converting those cameras as just being the security cameras, which can keep the bad actor out to actually becoming these video intelligence cameras, which can actually look at what’s happening in the frame and using AI, be able to produce a bunch of statistics and analytics which operators could use,

whether it’s the amount of time at the loading bay, unloading bay truck is waiting for, or are the right safety equipment being used while loading, unloading, and few different statistics around it. So we basically think of those cameras that you see at the corners of the building and then turning them from just security cameras to catch a bad actor into video intelligence.

Jamie Irvine:

Well, you know what? You’re absolutely right. I think with traditional video systems for security, you had to wait until something bad happened before you could react, and then you had to kind of isolate what time that bad thing happened and then you had to go watch that video.

Otherwise, the amount of data that was being captured by these cameras is so overwhelming. It’s like how do you possibly process all of that and turn it into anything meaningful? I think this is where AI is really interesting because it gives us the ability, it seems to me anyway that AI is very, very good at doing things that a lot of things humans can’t do. And one of that is processing large amounts of data.

Rish Gupta:

It’s processing large amounts of data. It is also about how do you get to the data? As you said that if you f found an incident, you had to find the exact time and you have to go look for it without any good AI filters, you’re just watching hours and hours of footage until you find that one thing. And with all systems, they still require you to use a thumb drive to a VHS like system, which is only on premise. So if you’re not at the loading bay, and very often your loading bay could be very far away from your headquarters.

Where the central office dispatch team is, we have one of our customers, they produce products which are time sensitive, which are produced, which can get spoiled, and sometimes their customers will receive a product batch from their trucks, which would let’s say not up to quality, and they want to prove that they did every right procedure during their loading, so they follow all the instructions.

So any loss of quality of product is not the transporters problem, it’s how the product came to them and they have to find this video, make sure it goes to the customer really fast. Using technology they’re able to really find these videos quickly and just like you would share a YouTube link, you’re able to create a secure link just for that enterprise customer and be able to share it.

So as transportation companies, they now have these modern tools which allows them to provide data not just for themselves, but also to their stakeholders outside of their immediate environment.

Jamie Irvine:

So is the technology only connected to external cameras or can that technology also be applied to monitor what employees are doing on their computers or other applications like that?

Rish Gupta:

Yeah, it’s both external and inside the building cameras. Hopefully, I don’t think it’s going to zoom into what each person is watching on their computer and hopefully nobody has to spy on their employees at that level of detail, but giving you basic idea that, okay, how many hours did the person spent in the break room versus how many hours did they actually spend on the floor?

Are there certain things within the floor which are obstructing safety pathways or exit pathways? Are they unblocked? Is there a loading bay which is full? These are the kind of information even from inside the internal cameras inside the offices can give you a lot of information.

Jamie Irvine:

I know sometimes people kind of jump to the idea that this is all about being punitive with employees, and maybe it’s just because I’ve been a business owner, I take a different perspective on it because the way I look at it is you want to take care of your employees and then your employees take care of your customers and then the customers takes care of the company and then the company takes care of the shareholders.

So if you have that mindset in your business, it’s not about being punitive per se, it’s more about figuring out ways to protect your employees, to make them more efficient, to make the company operate better on behalf of the customers. And this benefits everyone. And if you know your employees want long-term gainful employment, then that’s a good thing for them. It’s not all about just trying to catch them doing something they shouldn’t be doing.

Rish Gupta:

Absolutely. The simplest thing is there’s the whole Big Brother kind of image with the cameras that somebody’s watching you, that’s a big brother watching you. And to us by making the technology really simp to use, getting down to your mobile apps, making unlimited users, being able to access the camera footage, what our customers are able to do is give the camera access not just to the managers or somebody sitting in the headquarters, but to the line employees, to people on the front line.

And actually they start seeing the benefit from day one because let’s say you are at one loaded, they’re a hundred bays there. Now, if they want to know what’s happening on the 89th bay or the last bay, either they have to go walk, take a walk outside, see what’s happening, or call someone who might be there and see, hey, is that bay on?

Or if they have some kind of a dashboard, track it versus now they’re just able to pull up their phone even if they’re not in location and be like, oh yeah, we have these five bays free, or this truck was supposed to leave by PM and it’s actually left. I can see it in the camera. It’s no longer there. So this is good. It’s on the road. When employees have access to all the cameras they know what is being watched and why is it being watched, and that becomes a more transparent conversation versus if only somebody sitting in some ivory tower has access to these cameras.

Jamie Irvine:

So AI is kind of a very popular term now and it means a lot of different things. So how adaptive and how much machine learning is built into the application, and if so, how does that improve performance over time?

Rish Gupta:

Yeah, that’s a really good question. And without getting into too much of geek speak, I’ll give two simple analogies of what’s happening. One is, if you were to build a website 20 years ago, you would need to know how Linux servers work to spin up one yourself, you’ll have to understand different layers of code and write. Today, anybody can build a Shopify store or a website without knowing any programming language. There are tons of software tools available online. You can just quickly build it.

AI is going through the similar transformation era where 15 years ago just to tell whether an image is a cat or not a cat you needed a hundred R&D engineers in Google with the best supercomputers to work. Today if you want to build a cat classifier just looking at the image as a cat or not, you can download a program onto your Mac or onto your PC and run it within five minutes without knowing any coding, and you can start classifying images.

So AI has moved to this application there just from a pure R&D thing, which it was 10, 15 years ago, which means is that now companies like us can offer AI, which was been in R&D in different companies for 15 years, 20 years, but now it’s available for mass production and mass consumption. So people detection, vehicle detection, counting the number of people growing heat maps, how people are moving within a sector, are all your loading bays occupied, filling out idle times, all these things become very, very doable and it gets doable at a faster speed.

So it’s just like you saw once web technology became available, it improved so fast, you could build new and newer things on top new features. Similarly, AI will see this massive feature velocity, which is what we are looking at over the last year, we must have launched four or five new AI features and every year we’re going to launch that number of new features.

So then you’re looking at a 10 year period, you’re like, oh, you’ll have access to 40 and 50 AI models used. And the second thing that’s happening is the GPUs themselves, which is the processing unit for AI, is just like your CPUs for normal computing processes.

There’s GPUs for video and image AI processing, and those are getting better by 2x, so they’re doubling in capacity at the same price point every year. So every year a newer model can come, which can do twice as much as what the previous generation could do from a processing. So it’s like imagine your brain was getting double every year, and if you do it for five years, that means you’re about 30 times smarter, stronger, better than before.

So that’s what’s happening in the AI technology, which is why AI will no longer be this buzzword, which seemed like an R&D science fiction thing. But because of the application layer and finally the computing moving so fast, just like internet kind of spread across everywhere now everybody has a website, everybody has an online profile, whether you’re in LinkedIn as a person or as a company, you have some kind of a simple website, more complex technology company, you have more detailed websites.

Similarly, every company will start leveraging some AI. The depth of AI you’ll use will depend on how much your company needs it and the technical depth of your team. But AI is going to become ubiquitous in everyone’s head.

Jamie Irvine:

Let me kind of run three different locations that a transportation fleet may manage. So it may have a dispatch office, a logistics center. We might have a large warehouse, like you said, with a hundred bays for logistics and for warehousing. And then we also might have a repair shop for maintaining and repairing our vehicles. How customizable is your solution so that we could literally set up cameras in all three of those locations and gain valuable information that could then be used to improve the efficiency and the safety of the company?

Rish Gupta:

That’s a really good question. One of the advantages of what we have built and we have built with open architecture in mind, what I mean by that is it can connect to every brand of cameras. So if you already have some deployed cameras in your warehouse and your dispatch office and your repair center, you can use this existing cameras on a platform and start leveraging AI and modern video technologies on top of it.

Second, at a fundamental level, what we realized is there are two things which are moving in these spaces, whether it’s dispatch center loading, its people or vehicles. If we can understand people and vehicles context in a given image or a given video and then we can able to do a bunch of calculations on top of it, we can provide you unlimited amount of information which is contextual to the study.

So for example, in dispatch office, you can easily draw a zone around when dispatchers sit and see total people present time between 8:00 AM to 5:00 PM let’s say Monday to Friday, barring a lunch hour. And if you’re seeing certain dispatch officers working less hours versus more not being present because they were outside taking a smoke break or they take a longer lunch breaks on an average, you can find those insights and then be able to train your team to be more effective and proactive in these things.

Jamie Irvine:

Or not even something like that. It could also be something as simple as that particular dispatcher is spending 15% of their time training and helping someone else who’s struggling. So again, it’s not always punitive. It could just be that this person is doing a really great job, but they’re being taken away from their core responsibilities. And Rish, another application is like when you have a repair center and you’re bringing trucks in and out of bays and trailers, you would be able to then say, what kind of input and output are we getting in each bay?

How does that correlate to downtime and things of that nature? So I think there’s a lot of really interesting applications where you could, and from what it sounds like because it’s open source, there is a fair amount of flexibility on how you configure what you’re recording and how you’re going to then do a multi-factor analysis on that data. Is that correct?

Rish Gupta:

Yeah. And the other things we are doing, absolutely, and the other things we are doing is also building integrations to other systems that you might have where open platform, that’s also what I meant is other things that you might, other technologies that you might have in your building. So for example, if you have a badging system and you want to capture people as they badge and badge out from a video perspective, you can do that.

If there’s in the repair shop, if there’s certain machines, if it’s a POS system or something which is transacting saying, okay, this truck came in there, we are going to give it this item number, and when it gets transacted, this is how we can punch it out. Now you can integrate that into the video. So there’s a video evidence of the shape in which the truck came in, the shape in which the truck went out, and you can actually see things happening.

So there’s a lot of cool systems you can build once you have access to this visual data. And one of the key things we also realize is talking to a lot of our customers and seeing them use it is because as humans, we consume about 85% of our world through our eyes that once you have visual full spectrum of what’s happening in a room, you are actually able to see these insights.

As you said in the example of the dispatch officer, you may not see them in the seat, but then you see them talking to the other dispatcher, helping them, showing them the ropes, and you’re like, okay, this person is actually educating others and maybe there’s a potential to elevate them to a training role or do something else.

Jamie Irvine:

Or even maybe that’s an indication that there’s a weakness in your overall business systems and that you need to do better training so that all the dispatchers have the available information they need or the repair shop or out in the yard. Yeah, I think this is fantastic. So how would people get in contact with you and get started? How easy is it to get this whole system up and running?

Rish Gupta:

Yeah. First of all, it’s super simple. If you go to our website, spot.ai, you would be able to find all the details. Second, from a setup perspective, because we connect to existing cameras, we always ship as a box. You plug it into your internet connection and your power source and your up and running in 15 minutes. So it’s super easy to deploy. Anybody can deploy it. Our support team is happily to be on the call to help you go through it. The more interesting aspect is that we also offer, I think a two week free trial to any customer.

So the customer has no obligation to purchase, nothing that they’re signing, which holds them accountable or putting a card in file. They’re just getting, they’re like, Hey, we interested. This looks cool. We’ll send you a trial you tried out on one or two of your facilities. And then if it, you’re like, okay, I can see the advantages, and then we can talk about purchasing or anything else.

Jamie Irvine:

Well, thank you so much for coming on the show, Rish. For everyone who wants to check this out, go to spot.ai links will be in the show notes, and I really appreciate you taking some time to explain this technology and how it can impact safety for trucking operations. Thanks so much for being on the show.

Rish Gupta:

Thank you, Jamie. This was a wonderful chat. I appreciate you. Thank you.

Jamie Irvine:

Well, I think we have a better understanding about what it takes to make a fleet safe and how we can leverage new AI technology to increase safety and productivity. If you’d like to learn more, by all means, go to the show notes and click on the links for our guests so that you can learn more about their companies. Also, our show is supported by sponsors, so take some time to go to our sponsors websites and learn more about what they have to offer for you. If you haven’t already, make sure you go over and follow The Heavy-Duty Parts Report.

You can do that in three ways. So go to heavydutypartsreport.com, sign up to our email so you never miss out on any of our content. Or if you watch the video version of this show, go over to YouTube and subscribe to our show. And if you’re listening on a podcast wherever you get your podcast, you can follow The Heavy-Duty Parts Report for free. Thank you for supporting our show and listening today. And as always, Be Heavy-Duty.

Share this:


Receive a weekly email with links to the latest episodes.

You Won’t Believe How Much You Will Save.

Your Complete Shop Efficiency Partner.

Provide Your Customers with Complete Parts Visibility.

The Heart of the Independent Parts and Service Channel.

Your Ultimate Destination for Heavy-Duty Parts.