Listen to most recent episode of Road Signs

SEASON TWO

SEASON FIVE: EPISODE ONE

HOW WILL MACHINE LEARNING ENHANCE TRANSPORTATION SOFTWARE?

Trucking decisions have gone to the machines. The promises sound almost magical: supply your shippers with better freight visibility with one click, better manage your driver profiles and increase retention with this easy software, find the operational silver bullet without having to hire a data scientist. In fact, today, an algorithmic solution exists for almost every trucking and logistic problem. And, for those big data set problems not yet covered, there’s capital available. But, Seth Clevenger wonders, how well are the machines learning at scale? Will these technologies really continue to provide today’s trucking and logistics managers with better decision and support solutions for real-world problems tomorrow? Or, are the profitability promises about to fall flat? To get some answers, our host travels to the 2019 McLeod Software User Conference and sits with tech expert and company CEO, Tom McLeod, who helps to put a human explanation to the work of the machines. Listen in, before RoadSigns automatically recommends a different podcast episode.

FEATURED GUESTS

Tom McLeod is the founder and CEO of McLeod Software, a developer of transportation management systems for motor carriers and freight brokers.

Tom Curee is the VP of Strategic Development for Kingsgate Logistics and is a thought leader in leveraging new technologies in the transportation industry. He has spent over 13 years in the transportation industry, working with both brokerages and carriers. Tom works with the Owning Partners of Kingsgate, driving innovation and strategy to prepare for the future of our industry.

Tom McLeod

Tom Curee

EP. 1

Brought to you by:

Episode Transcript

Dan Ronan: From Transport Topics in Washington, D.C., this is RoadSigns. Here is your host, Seth Clevenger.

Seth Clevenger: Thank you for listening to RoadSigns, the podcast series from Transport Topics that explores the trends and technologies that are shaping the future of trucking. In this episode, we're going to examine the transportation management software that lies at the core of most trucking and logistics operations. These systems have evolved from basic dispatch software to the more sophisticated operations and business intelligence platforms that we see today. But now the trucking industry is facing a time of significant change, with shippers demanding constant visibility of their freight and transit and an influx of capital fueling technology development. At the same time, fleets are under increasing pressure to improve driver retention and recruitment. So how will dispatch and operation software evolve to meet these challenges? We'll set out to answer that question in this episode.

For starters, some of the biggest trends in technology development have been making their way into transportation. Data science, artificial intelligence and machine learning are beginning to unlock more actionable information and provide better decision support for trucking and logistics managers. I spent some time considering these possibilities at McLeod Software's 2019 User Conference in Denver, where I spoke with the company's CEO, Tom McLeod. I also sat down with Tom Curee, vice president of strategic development at Kingsgate Logistics, a freight broker that is already doing some very interesting things with AI today. Let's go ahead and get into those interviews.

Seth Clevenger: We're here in Denver at McLeod Software's 2019 user conference, and we're very pleased to welcome our illustrious host of the event's CEO, Tom McLeod. So thank you very much for taking time out to join us, Tom.

Tom McLeod: Thank you, Seth. And thanks for joining us here in Denver.

Seth Clevenger: Yeah. First off, you know, it's always great to be here at this conference. You know, I always come away with a lot of great insights and ideas.

And, you know, I'll just mention that it's pretty remarkable to see how much it's grown really over the last handful years. So thanks again for having us here.

Tom McLeod: Well, we've had great customers and we've had great support from allied suppliers and the industry, and they help us to put on a really great meeting.

Seth Clevenger: So one of the big themes that you addressed during opening session was the sense that there's this transformational change occurring in the industry. You know, you mentioned the advance of onboard technology, information technology and changing freight patterns, really all coming together. But, you know, I'll let you speak to that yourself. You know, what are the main reasons you feel trucking is in the midst of a major change for the industry?

Tom McLeod: Well, in some ways, the freight industry and all industries are undergoing some change and some transformation when the change is happening pretty slow lately, almost call that status quo and is happening so slowly that changes can happen without us realizing it. But the pace of change has certainly accelerated, especially over the last few years, and especially as you have so many ideas being funded from outside the industry and all kinds of solutions and mainly technology solutions appearing as options and capabilities that have an impact on the way freight is sourced, the way freight is covered, matched up, the way it's billed and paid. And so the accelerated pace of change and the new solutions coming out of the existing carriers is, I think, unprecedented.

Seth Clevenger: No doubt as busy times and there's a lot to watch and monitor as we see all this technology and all these, you know, like to your point, new players jump into the industry and attempt to disrupt it. But, you know, I like your comment that change can also be an opportunity. You know, it doesn't have to be a threat. So I wanted to go into that a little bit more. You know, in your view, how can motor carriers and 3PLs stay ahead of the transformation changes that we're seeing and benefit from them?

Tom McLeod: Well, certainly change creates challenges and don't want to minimize that. But change is definitely gonna create opportunities especially if you have the right point of view and you're looking for the opportunity within a change rather than the challenge. You don't have to capitalize perfectly on the change. But if you can see an opening and be the first one in the door, you can move ahead of competition and really reap some great benefits.

Seth Clevenger: And you also spoke about artificial intelligence and machine learning. You know, of course, you know, this is one of the huge trends that we see in software development. But how are you applying it today at McLeod and what might you be able to do in the future through AI and machine learning?

Tom McLeod: Well, where I mentioned the two main areas where I believe artificial intelligence and data science. And then machine learning can have the best positive impact on transportation operations, one is in making pricing decisions, and that's a strategic contract pricing as well as spot pricing decisions. The secret to that is analyzing all the relevant data and then trucking and transportation, the data that you've got access to. A lot of the relevant data is just a company's own internal data. And so there's not the millions and millions of data points that you might have if you're analyzing, say, consumer purchasing habits. And so the problem's actually, from that standpoint, somewhat easier to attack. But properly bringing all the relevant data helps you come back with good recommendations. And it can really give an even midsize to large carrier some strategic advantage in the transportation marketplace. The second area where I felt like this can have the biggest payoff is in the matching the right truck to the right load in less time. So you’re seeing productivity increase, as seen better utilization of a fleet that translates into profitability. Profitability lets you invest in other ways to move ahead of your competition or serve your customers better. And that translates into staying power in the marketplace.

Seth Clevenger: Yeah. Another topic I want to touch on is digital freight matching. You know, this has been a pretty big topic in the industry in recent years with the emergence of some of the digital freight brokers like Convoy, Uber Freight, Transfix. But how do you think about this trend and really how are you participating in it?

Tom McLeod: Well, you name off three freight brokers. Yeah, that's right. They have their own internal technology, but those companies really have closed systems. They put their freight out on their internal network and then they've got carriers, a lot of times individual lease operators that sign up and that say the freight is visible to, you know, we're participating in several versions of early more open platforms where visibility is more widely available and where there's more networking opportunity. We think that many of the brokerage operations are gonna do just fine. You know, in that environment where they're looking to match up freight capacity and more or more the carriers are getting comfortable about sharing capacity information with their brokers as they're competing against, again, the closed systems, companies that are really bent on maximizing margins, even though they may be, you know, offering temporary loss, loss leader pricing to try and gain access to capacity.

Seth Clevenger: Sure. And, you know, one of the questions that comes up when we look to automate freight transactions, and to streamline some of the processes and, you know, make some of the use, the standard procedures, the grunt work simpler. We have this question of how do you balance automation with human control and a human touch? You know, what is the right way to go about business process automation in an industry like trucking, especially freight transactions while still balancing the need for, you know, the to protect the relationship and build a relationship between a trusted relationship between a shipper and a carrier, 3PL ... all these parties.

Seth Clevenger: Well, if the routine transactions can be automated, it actually frees individuals up within the transportation company to give more time and attention to the exception transactions. It should also allow more time for face-to-face contact with the customers and in building the relationship. The relationships are still important in the transportation business because when the shipper trusts you with their freight, that's their livelihood. They've got things that are very important that make it to the destination on time because that's the lifeblood of their business. It's a little bit different, although pricing is always a factor.

Tom McLeod: You know, one of the things that was very tough on the electronics stores, remember Circuit City?

Well, you know, people could go buy look at an article they wanted to purchase, compare prices and brands in the store and then go back and order it online. Very little loyalty translated there, a little bit different in the freight business. Again, because that long-term relationship, you're not just buying one. You're not just transacting one load this year. You get most of the shippers are doing many loads on a weekly basis. And again, very important that they maintain that access to the capacity during good times and bad. That said, there's obviously still price sensitivity based on available capacity in the marketplace. And, you know, certainly anybody with enough price differential can be swayed to go on, go another way.

Seth Clevenger: You know, another thread that we hear about the business environment right now is that super expectations for their carriers and brokers are rising. You know, they're placing more demands on transportation companies and they want more freight visibility. So how do you see that trend, Tom? And what can transportation companies do to meet those expectations moving forward?

Tom McLeod: Well, especially the higher volume shippers, the carriers and the transportation managers, the freight brokers are, you know, sometimes somewhat at the mercy of the big shippers. And frankly, a lot of times the shipping department at a large corporation are not all that sophisticated. Most of the large manufacturers and production companies have their best people making this stuff and selling it to their customers. So sometimes the request being made by the shipper representatives and the traffic department are simply blanket requests and can be very difficult to supply with or comply with being able to have good ongoing relationships with the shipper where you're helping to educate them on what's happening and what's possible.

And then listening to them to hear about the things that would be truly beneficial really takes effort. It really takes face time. And when a shipper makes an unreasonable request that strong communication if you've got that pipeline in place, then you can a lot of times work to come up with some more reasonable complaints. At the same time, the shippers definitely want more visibility because they want to provide that to their customer. Right.

And you know, constant visibility, sometimes of all shipments has become almost table stakes. Right. Price of entry and you know, certain markets. And I'm sure that's going to expand, especially as it's become easier and easier to know to get that visibility. So I would expect that trend to continue.

Seth Clevenger: Sure. And shifting gears a little bit, I did want to touch on one of the more interesting new features that McLeod has rolled out recently. It's your Driver Choice for LoadMaster. And this is enabling drivers to create preferences for the types of loads they like to deliver. So what do you think? I mean, do you think the industry is going to move more in this direction of giving drivers a little bit more, say, exactly the types of jobs are taking in a part of an effort to boost driver satisfaction and retention, which, of course, is always a top issue for trucking companies?

Tom McLeod: Oh, yes. We know from the reaction we've gotten from our customers and prospective customers and in other industry segments, we think it's a very powerful thing to engage your drivers, help them not only feel like they're part of the process, but help them to actually be a part of the process. And I think the companies are pleasantly surprised that the drivers don't just go crazy with a lot of their requests. They're reasonable people. They understand the company has to make it. But for them to be consulted and have some input on their load assignments, especially your experienced people. This pays him a great compliment and can engender some great loyalty there.

Seth Clevenger: That will be interesting to watch that moving forward. Another big topic in recent years has been blockchain, certainly in the world of, you know, the technology world. It's just been, you know, everywhere. That conversation, of course, is carried over into transportation. And we've started to explore how this shared ledger technology could really help to streamline certain freight transactions and certain applications in our world, and McLeod, of course, as a founding member of the bloc's main transport alliance, which has been putting in a lot of the work on the front end of this. But over the past year or so, I think it's certainly safe to say that, you know, the hype has died down considerably. But that doesn't mean that there aren't still real opportunities and real applications on the horizon. So what do you think about the potential for blockchain today? Where do you see it?

Tom McLeod: Well, the potential for the technology is the same as it always was. And it's very common for a new technology to be overhyped when it first appears. And then for there to be disillusionment when it doesn't immediately deliver. And this is a technology that will require work, especially within the supply chain standards, to get things into a common format. The shipping community, there's a couple of large shippers with private blockchain projects underway. The last thing we need to see in the transportation industry is for 50 of the largest shippers to come out with their own format. And the carriers then asked to comply with 50 different ways of communicating the same information. If we can agree on a set of standards, then a lot of the visibility that the shippers are looking for today will be easy to supply and it'll be we'll be able to supply it through direct networks without going through third parties and aggregators where sometimes there's a delay on getting the information. Or there can even be, you know, accuracy issues with some of the current providers, so the technology shows great promise. We've got 600 members currently. That does include some members of the shipping community. So we're optimistic that they use cases and the standard formats are going to be useful in getting them implemented across the board will be the neat trick. If we pull it off, it'll be very beneficial for the industry as a whole.

Seth Clevenger: Now, Tom, you've been developing software for trucking operations for quite a while now. You know, 36 years is the estimate that you provided since you have this perspective. You know, I want to ask you just how much you've seen technology change this industry, how much this industry has involved, evolved through the use of technology over the years and just how much you think it will continue to do so from when you started to where we are now and where we're going?

Tom McLeod: Well, certainly the capability, the systems, the change there has been amazingly remarkable. Sometimes I say that our original system would print a freight bill and print a check and that's it. We were replacing typewriters, you know, in the offices of the trucking companies and just simply reducing manual labor, to that extent. As we got into operations and planning, we began to see better utilization of the equipment, the carriers asking for more. You know, in a competitive situation, the programming technologies got better, allowing us to develop more quickly. But these days, even a version one product has to be robust, you know, for it to come out the door. And the trucking companies’ requirements from the shippers is light-years away from where we were at the beginning. We've had great input from our customer base. The great thing about working with trucking companies is they'll tell you what they want. And if you can scramble and get there, you know, they'll be loyal and continue to give you good input.

And thankfully, we've had a great group of customers that has kept us pointed in the right direction now for a long time.

Seth Clevenger: And you see it continuing in the future. I mean, just more and more change through the use of technology. I imagine that, you know, in 36 years this industry will look quite different yet again.

Tom McLeod: Oh, I think so. Yeah. There's so many new technologies, you know, really many of the newer technologies in the truck this morning. I mentioned, you know, lane departure warnings, collision avoidance, adaptive cruise control. Those are actually software based. You know, a lot of the technology is on the trucks for EPA compliance over the last 15 years and has been at the hardware level. But the behavior of the truck, even the recording of the images and knowing when to capture, you know, an image for future reference is based on software. So a lot of those products are things we interface or, you know, capture information for. And so a connected system that helps evaluate and provide the right recommendations to the user and to the trucking company management is crucial to the efficient operation of the trucking companies today.

Seth Clevenger: It'll be very fascinating to see how this all evolves in the years and decades ahead. But, you know, this has been a great conversation. I think we're a good stopping point here.

So we'll let you get back to the rest of the conference. Thank you so much for joining us, Tom.

---

Seth Clevenger: We're here in Denver in McLeod Software's 2019 User Conference, and we're very happy to welcome Tom Curee, who's vice president of strategic development for Kingsgate Logistics, a freight broker based out of Westchester, Ohio. Thanks for joining us, Tom.

Tom Curee: Glad to be here. Thanks.

Seth Clevenger: So we're gonna talk about AI and machine learning and you're a real expert on this topic. But let's just start with the basics. What do we really mean by AI? You know, this isn't Skynet from Terminator, right? So what are we talking about?

Tom Curee: So it's kind of a funny journey, because even myself, as we talk about artificial intelligence, you know, five years ago, I wasn't a guy who really would have sat here with you and told you, “Seth ... well, this is what AI is all about.” But really, it's pretty simple. AI is really just using technology to make decisions that humans typically would make. Now, this isn't talking about having robots go get your groceries and do everything else. It's right that we're seeing today. But it's about really just leveraging this technology so that it can make decisions that you're already routinely making all day long.

Seth Clevenger: And this concept of AI isn't new. You know, even early calculators were thought of as artificial intelligence at the time. But now this has been a really huge topic in the past handful of years throughout the technology world. So why have we seen this resurgence in AI development?

Tom Curee: Yeah, I like I like as people talk about buzzwords and technology and they think that AI is brand new thing that no one's ever heard of. But really, if you look at it, AI has been around since the ‘50s. This isn't a new concept. It isn't a new idea. But what has changed about the idea is the use of big data. You know, six, seven, eight years ago, we were all hum and hey, big data. It's where it's at, big data. But the reality is, is once we got all this big data, we didn't have a clue what to do with it. There was just so much of it. And so we began saying, okay, how do we use this? And all of a sudden we go back into the playbook and we go, wait a minute, what about this artificial intelligence stuff? We were doing something with it. Now we've got more data than we know what to do with. Maybe we need to revisit this.

Seth Clevenger: Yeah. How are AI and machine learning being applied in the transportation industry today? Can you provide some quick examples of what you're doing at Kingsgate?

Tom Curee: Yes. So AI machine learning, you know, we're primarily working in that space. You know, I dream of being able to play with deep learning, as I know we talked a little bit about. But as you really think about what we're doing today, we're using it for carrier selection primarily right now. So, you know, McLeod has done a nice job of building their product top match where we are, we are feeding in data into McLeod to decide, hey, these are the type of carriers that we typically like based off of the shipment, based off of the time that we have to cover the shipment based off of our previous experiences and a number of different factors that we can feed into that algorithm to then tell us what is the chance of success for this is at 92% is at 88%. And then we're leveraging that data to then learn from it and say, OK, typically we're actually picking these guys that weren't 90%, they were 80% or what did we like about them more? And we're tweaking that algorithm as we go.

Seth Clevenger: And, you know, one way of thinking about AI and machine learning is moving from a predictive standpoint to more of a prescriptive approach. So now software can actually tell you, you know, which carrier is the best fit for a particular load, you know, how is that changing the way that your company is making business decisions?

Tom Curee: So it's really interesting. You know, we find ourself in this new era of leveraging technology and having to do it so quickly. And so it changes the way we've typically made decisions. You know, a lot of times decisions in companies our size, they get held to 5, 10 people that maybe they've got all the experience. Maybe they've had the they've seen what it was like in 2008. They, you know, they've seen these experiences. And what we're finding now is we're able to arm brand new employees, new employees with technology and data so that they can actually make those same decisions that people maybe, maybe couldn't have made five years ago. And so we're able to get them up to speed a lot quicker.

Seth Clevenger: Great. And, you know, another application I think is really interesting is recruiting. So how are we using AI to find and hire the right job candidates?

Tom Curee: Yes. So recruiting was always a big opportunity for us in the freight brokerage space we've got. We typically are looking for a lot of people for sales or operations. And even as we've expanded our technology team building out that group. And so we are actually using a product that allows us to kind of refine what we really look for in a candidate of what maybe has their previous jobs been and maybe what type of skill sets do these people have or maybe what are some of their affiliations that they're a part of or are tenure, whatever it may be. And as we feed this in the system, what it does is it produces a number of candidates, a kind of a pool of candidates that we then review in our recruiting team will go in and they'll look at five candidates and they'll tell the system, hey, is this person a fit? If they are great, they funnel it over into the pipeline. If it's not a fit, it prompts them and says “Hey, why is this person not a fit?” And all of that data is being fed back into the algorithm to get smarter and learn better so that eventually the candidates that we see are typically candidates that we want to have conversations with. And so it slims down a lot of the time that we spend sourcing and just digging through resumes.

Seth Clevenger: Absolutely. And you guys are 3PL, of course, but you can really take that further and think about how that same approach could be used to hire the right drivers and reduce driver turnover at a motor carrier. What kind of potential do you see there to help solve? You know really well the industry's biggest problems.

Tom Curee: Yeah, I think I think there's significant opportunity there. You know, Seth, I spent time at a refrigerated carrier. We had 200 owner-operators. And when I first came to the company, we had turnover of 133%. And so imagine just the sheer amount of recruiters that we had to have and the candidates that would come through and not to mention just the awful driver experience. I mean, you know, it's hard for guys to start building relationships with people that they're driving with, kind of know these companions on the road for that person to just be gone. And so when you start thinking about the attributes of what you're really looking for from their experience, I think you talk about a technology like that and you can say you can reduce that turnover by 20, 30%. That's a significant play, not just for the company, but also for the experience for the other drivers that are there.

Seth Clevenger: No doubt about that. And let's just look ahead a little bit, and I want to get your thoughts on what you think the future might hold for AI. I mean, are we going to get to that point where we do move into deep learning and we're, you know, software is really training itself to, you know, improve itself on even more data?

Tom Curee: Yeah, I absolutely think and I definitely think that we're gonna get into more deep learning. You know, when we talk about deep learning, I think what you'll find is because of the sheer amount of data and the investment that people are going to make in the technology, you're gonna find a whole lot more people playing in that game. And that's when you start taking I mean, we we've got so many data sets coming at us right now, and that is where you're going to be able to say, OK, here's how these data sets can be associated together. And here's what I want to do. I want you to look at every piece of data that could be connected and determine based off of every data set. What's going to be the best opportunity for us going forward?

Seth Clevenger: Sure. So we see all those offered opportunity. But, you know, I do want to ask you why Kingsgate Logistics really decided to be at the forefront of this and really start to put this technology to use, because certainly not everybody in the industry is.

Tom Curee: Yeah. So we kind of started on a path last year working with our ownership team. And we've got a great, great company, just a phenomenal ownership team who are a second-generation family-owned business. And as they've looked at the company and where they see it going for the next 10, 15 years, we're looking at that third-generation family who's gonna come in. They're going to continue to be more active. We have three members of the third generation who are actively working at the company. And they said, you know, we want to make sure that our business is thriving. We don't want to give. We don't want to have this business for our family that is just kind of on the downslide. And so we started looking and say, what does the future look like for our industry? And as we continue to look, you know, we found just such a significant need to be a player in technology. You talk about digital freight brokerage, digital freight matching, pretty much everything that you could name that starts with the word digital. And we said, OK, well, wait a minute. We can't be the traditional brokerage that we've always been and not play into this digital space. And so we really saw that investment in our ownership team just this year. I mean, they are making a seven-figure investment in technology, which is like nothing we've ever done in the past. And so we've got the vision. And along with that, our customers and our prospects, they're jumping on for the ride. And as we've seen buy-in there, we've felt confident that we're going down the right path.

Seth Clevenger: Well, I'll be really exciting to watch in the years ahead. But just a final question for you, Tom. You know, just how much opportunity is still out there for the transportation industry to really do more with AI and machine learning in the years ahead. How much further is this going to take the industry?

Tom Curee: I think I think the opportunities is incredible. Most of the AI and machine learning that we are leveraging today, it didn't, it wasn't birthed in the transportation industry. It was birthed in other industries. Birthed healthcare and is built in marketing or whatever it may be. And we've been able to grab it, adopt it, you know, kind of tweak it a bit to make it work for transportation. What I think you'll find is you're going to find some technology providers who are gonna step up to the plate and they're going to start leveraging some of this same technology and it's going to change the way that we do business going forward.

Seth Clevenger: This has been a really great conversation, Tom. It was really a privilege to have you here. And thank you so much for for sharing your insights with us.

Before we wrap up, let's take a moment to revisit our original question: How will dispatch and operation software evolve to meet the challenges of the future? As we've heard in this podcast, AI and machine learning will open up new possibilities for software development. This will create a wide range of benefits for business management, including smarter pricing, better freight matching decisions and improved hiring and retention. The transportation industry is already capturing a vast amount of data. Now developers are looking for ways to better utilize it through improved analytics. Expect routine business processes to become increasingly automated, which will free up workers to focus more on customer service and responding to the unexpected events that are inevitable in transportation. While the chatter about blockchain has subsided, it's still on the horizon and will likely begin to realize its potential in the years ahead. And as AI continues to advance, the transportation industry will move toward deep learning. Software will begin to train itself to get better and improve through experience. The sky's the limit on how much more efficiency and decision support we can potentially unlock from all the data that we're collecting.

RoadSigns will return soon with more episodes exploring transparency and business intelligence in the transportation industry. Until then, I'm Seth Clevenger. Thank you for listening.