What better rail ETAs mean for grain shipments

Steven Kilger: Hello, my name is Steven Kilger. I'm the managing editor of Feed & Grain magazine and the host of the Feed & Grain podcast. I want to thank you so much for joining me today as we dive deep into the issues affecting the feed manufacturing, grain handling, and allied industries.

Today's episode is brought to you by the Binwip from Pneumat Systems. The powerful dual-impact Binwip removes the toughest buildup and blockages in industrial storage silos without hazardous silo entry. Learn more today at binwip.com.

Today my guest is Andy Adams, senior solutions engineer at TransmetriQ, a subsidiary of Railinc. TransmetriQ aggregates rail data from across North America to help grain elevators and feed manufacturers solve one of their biggest headaches: knowing when their shipments will actually arrive.

Their platform delivers accurate ETAs so you can plan receiving, staffing, and operations more efficiently, turning rail transportation from a guessing game into a predictable part of your supply chain. I hope you enjoy the interview.

Andy, I’m really excited to talk to you more about transportation in the grain industry. But before we do all that, can you tell me a little bit more about yourself and your place in the industry?

Andy Adams: Sure. Andy Adams. My technical title is senior solutions engineer here at TransmetriQ. What I normally tell customers is that's really a catch-all title where I spend the majority of my job helping integrate our products, our platform, and help them use that platform and integrate that into their business. I'm also a part-time data scientist creating reports and playing with our data and coming up with new ideas. Then I also do a fair amount of teaching, both internally and externally for customers on how the railroads operate and think about the world.

I've often joked internally, Steven, even employees who join Railinc become part of the industry and TransmetriQ. The only thing they know about railroads when they join is generally the train blocks the road on them sometimes, and they have no idea why.

Kilger: Yeah, it's funny for our industry too, which is, you know, we're covering grain elevators, feed manufacturers, allied stuff. Before I started in the industry, I was just like, oh, what are these big concrete buildings out in the middle of nowhere? It'd be really good for like a zombie apocalypse or something. It seems like you got good sight lines and things, but it's interesting, isn't it, how little of the overall system we can do, even if we're related to it. If it's not directly our jobs, we just don't gain a lot of information about it most of the time.

Adams: Railinc, if you don't know, is a wholly owned subsidiary of the Association of American Railroads, and what we do mainly for them are back-end IT services, shared IT services that all six class ones and the 600 short lines need to use. So we become an industry sort of repository for those services. Like we manage the Umler file, which is the characteristics of all the rail cars. We manage the industry reference files. We manage a lot of industry car hire and interline settlement transactions for the industry.

Because we do that, we get about 14 million movement events a day on rail cars and containers from all across North America. And what TransmetriQ does, we are a brand of Railinc, and because Railinc does those back-end services, the railroads allow us to aggregate all of that data in a single place. With the proper permissions, customers, even in your space, and a lot of customers in your space come to us so they can come to one place to get all of that data and not 600-plus connections to try to find that data.

Even within the customer base that I would deal with that you're interested in, there sometimes is a lack of rail knowledge and how railroads actually operate and think about the world, Steven.

Kilger: Yeah, I would say definitely so. Well, because you mentioned like just how big the industry is, how many different companies are involved, how many different people you're dealing with. It's a lot to keep track, and especially grain industry, where we're shipping more product than ever before. And we're doing it with less people than ever before. So really, they need that data. They need that information from someone else to just stay up to date. Right?

Adams: We deal with all the parts of your supply chain on rail, whether it be the large agri-producers that are shipping 125-car unit trains around the country or exporting that out to foreign countries. We see all of that unit train traffic. We deal with the folks that are on the production side that are getting cars from those large producers of grains and producing downstream and consumer package goods out of those grains. We deal with the intermodal side of the house, any of those products that go export in containers. We deal with all of that and all of the entities involved in that supply chain.

Kilger: Yeah, I'm looking forward to nailing down some of the details too in those individual segments because it's interesting to me personally as someone who doesn't have to deal with the day-to-day and also just it's so much for our readers to keep track of on any given year. Grain in this country, I believe, is still moved by inland waterways. Next biggest chunk is by railroads, which move it from destination either across country or out to export. Can you tell me a little bit more about what the challenges grain shippers are facing today when they're moving their products, especially from rail to ocean for export or final facility?

Adams: All of that export goes two different ways. There is the 125-car unit trains that are getting dumped into cargo ships at the ports and sailed over. That's probably the vast majority of that business, especially in Canada. In Canada, I think one of the biggest challenges year to year has been rail car supply, has been adequate rail car supply to get the harvest out to port, particularly in Vancouver. And I know Canadian National and Canadian Pacific have to submit detailed plans early in the year to Transport Canada about how they're going to handle that car supply and ensure there's adequate car supply so the product isn't piling up in the elevators and not making it out to port in that regard.

Then they also need to know ETA times, which is always big in the rail industry. ETA stands for estimated time of arrival, and that's no matter what segment of railroad you're dealing with or what part of that railroad supply chain you're a part of, that estimated time of arrival is always critical.

What I always hear, Steven, is people ask for accurate ETAs, and accurate becomes one of those relative terms. It's like when someone says to you, I need it soon. That's sort of that same relative concept. So they need accurate ETAs to know if that product is going to get there in time to make a scheduled sailing on the export side.

Anything that goes containerized, because I know a lot more of this is moving in containers out the ports, they have that challenge of knowing is the container going to get there in time to make that scheduled sailing? Is it going to need to sit in that regard?

That's probably the biggest challenge they deal with for us on those unit trains that are going out export. Our visibility kind of ends when the rail car gets to the port, usually some large terminal. All of the big grain shippers, the G3s, the ADMs, the Cargills, whoever it is, they all maintain terminals inside the port to deal with that.

Then on the containerized side, on the intermodal side, they need to know if it's going to get there to make that sailing. And then probably the biggest black hole that I see in people's supply chain is when the container leaves the train and makes the boat. There's not a ton of information about, did my container make that ship in time?

Kilger: When I talk to people in the industry about what frustrates them when they work with railroads, a lot of it is kind of that lack of transparency, knowledge. I mean, the information's there. It's harder to get of when exactly things are coming and when things are going to be delivered, that kind of information, especially when you're at harvest time and you got to move stuff out or you got to get things ready for harvest and you need that exact dates when you're scheduling people, all that stuff.

So can you tell me a little bit more about how you guys are doing intermodal tracking and what kind of effects you're seeing with grain shippers who are using your system?

Adams: Intermodal tracing works like this. When a container is loaded or packed or whichever of those phrases you want to use, some of the shippers have their own intermodal facilities now, like I believe ADM does in Decatur. When they load those and get them on the train, when the container gets placed on the train, the railroad creates an association between the containers loaded on the intermodal flat car.

Then as the train moves, Steven, there are things called AEI tags, automated equipment identification tags on each of the flat cars. And so as those cars move, so the train's moving from Decatur to Los Angeles, just as an example, as the car passes various AEI scanners as the train moves, the railroad creates an event, what we refer to as an event on both the flat car and the containers.

As the container moves, you get various milestones that update you on where your container is in that regard. And you also get that critical estimated time of arrival. That's kind of how the tracing in intermodal works in the updates and those reports.

The second part of your question was?

Kilger: What advantages have you seen? What have you heard from grain shippers who have adopted your particular system? And what kind of benefits are they seeing from it?

Adams: The primary benefit is that ability to just have one place to get the data. We work with one of the major grain shippers, and they had that debate internally about do they just source the information directly and manage multiple data feeds, or do they go to a place where they can get all that information from? Because of our place in the industry, we see every container move that's on rail. And then the shippers tell us.

So what we also have for TransmetriQ is our TransmetriQ platform. We can help you manage all of that within our shipment platform that shows you where all your containers are, what the ETAs are on them, what the last event is on them to help you create metrics around when containers are supposed to arrive. You can bucket those days into 24-, 48-hour periods.

What it allows them to do is have an accurate look on where the container is and where it moves. And so if they're dealing internally, like even on an import or a domestic move, we also have the information on when the truck goes out the gate, when the container was unloaded, when it goes out the gate, all of that information. And it gives them an intuitive platform to build insights around where are my shipments, when are they going to arrive, what shipments do I have problems with, what shipments are delayed. And it gives them one place to build alerts and insights into that.

Kilger: That sounds really useful because sometimes, especially smaller co-ops, smaller grain companies, they have that old farmer spirit in them where they maybe have legacy systems that they've just kind of been patching together into one, or they want to do it themselves. Can you just explain a little bit what it's like to use the TransmetriQ approach? Is it all online? Do you get logins? How is it better for company overall to just take this off their plate and get this information through you guys?

Adams: There's a couple of different advantages. One of them you just mentioned is we're all cloud-based. There's no software integration, there's no software implementation on your side to use the platform. Everything is cloud-based because we see all of the railway bills. That waybill is that legal document for moving the container on rail that gets built out of the bills of lading. And because we get all of that railway bill information, we can find your cars for you. So you don't have to interface with us and say for every car or container you want to trace, go find this one, go find that one.

We can do that automatically. We pick those up at the start of what we refer to as a trip. We watch the rail car or the container make the trip. When it ends its trip, we drop it automatically for you unless you own or lease the rail car. Nobody really owns or leases containers. Few do today. And so it can be largely a hands-off approach for you where we're finding that information for you, putting it into dashboards and insights you define, and you can find your information.

It's all cloud-based. We do deal with a lot of smaller customers that like that feature, where they have fewer resources to devote towards, like you said, managing this or fewer dollars to invest and managing it all in-house for themselves. We've replaced a lot of spreadsheets over the years, even people still trying to manage their business on spreadsheets.

What I see very often out of our customer base is what you refer to as employee planning. A lot of the elevators especially need to know when those trains are going to show up so they can do their planning, their production planning, their resource planning for unloading the train and dealing with all of that.

Kilger: That I would imagine with staffs being so tight and just not having a lot of people, it's incredibly important that if you schedule everybody because you think a train's coming in on Friday, but it doesn't get there till Saturday, well, then you have to send people home and you have to do all that stuff and try to get a bag. Accuracy is the most important thing, right? They just want to be able to plan more than anything else. And it seems like you guys are really investing that with things like ETA prediction.

Tell me a little bit about how that works, because that seems like it would be really useful for being able to set up staff changes and when you're taking in grain and all that kind of stuff.

Adams: There are two different ETA processes that we employ. The first one is Railinc themselves plays a role in industry estimated time of interchange, estimated time of arrival, and that each of the roads puts in there what we refer to as a leg of the trip. Railroad A will put in the origin to an interchange. They would send that to us. We will electronically send that to railroad B. They would put in their piece and out it goes as a final ETA guess. And those could be dynamic. Those can update one to many times in the trip of a car.

And if Railinc doesn't see one of the railroads put in their portion of the trip, we have what we refer to as an elapsed time matrix that makes our best guess based on the last 90 days of history for that product type in that lane for what we see. That's the standard sort of industry ETA process.

A few years ago, TransmetriQ went down the road and we created what we refer to as an advanced ETA, where we're using machine learning in our huge data lake that we have internally where the customer provides us the lanes they're most interested in, that lane being the origin to destination. We do a training set based on the observations that we've seen in that lane. And then if a customer chooses to employ that, we will provide them with an advanced ETA around when we think the car is going to get there.

The results of that is largely dependent on numbers of observations in the lane. The more observations you have in data science, the better you are at predicting what goes on. But in general, we're about 20 or 30 percent better than the industry process within those lanes at being accurate.

Kilger: That's a big difference when you're talking about an industry that kind of lives on the margins.

Adams: With that size data, though, it can be a little hit and miss at times, honestly, to be frank about it. And it really depends on the consistency of the lane in numbers of observation.

For intermodal, it's a little better because intermodal in general can't just run on any track because of the heights of double-stacking containers on the trains. Years ago, they went through the process of lowering tracks underneath overpasses, enlarging tunnels. You just saw that project complete in Baltimore with the Howard Street tunnel where they're going to be able to start double stacking containers out of the port of Baltimore. And so intermodal has to run on those defined routes by and large. It's a lot easier to predict intermodal traffic, especially when you base it on train type.

There are different priority trains in intermodal where the domestic intermodal that runs solely within North America gets a little more priority, a little more timely than the international intermodal that runs.

Kilger: Interesting information, and it sounds like it could help a lot of facilities. So let's say someone out there is interested, they like the opportunity, they want to talk to you more about it. What is the process of reaching out and talking to you guys and getting a system like this set up at a grain elevator, a grain exporter?

Adams: There's a couple of different methods on reaching out. You can go to railinc.com or transmetriq.com, and there are links there where you can express interest. Reach out to us in general. One of our business development representatives will then reach back out to you.

What normally happens is they'll talk to you about what your interest is. Then they'll set up a demo of our services where you would get me or one of my teammates and we would pull up our platform, show you how that all works, how it would benefit you, try to get a very in-depth understanding of what your challenges are, what you're trying to do and how you're trying to improve your business process.

We will walk you through how we could solve that with our platform tool or with some reports outside of there. We end up doing a lot of utilization and cycle time reports on car fleets, fleets of cars or lane performance for your performance. Then they would work with you.

If you like that demo and you're interested, typically what happens is we would enter into a pilot with you for 30, 60, 90 days that we would set everything up for you, let you try it out, let you utilize it, let you ask me all the questions you want to ask about solving this, and then hopefully we move forward with you.

Kilger: I would hold out for you to come. They would tell me another teammate's coming. I'm like, no, no thanks, Andy, for me.

Well, thank you so much for talking to me, Andy. I really appreciate it today. It was great. I love the information. It's so nice to see more in-depth look in how this part of our supply chain really works.

Adams: Sure, my pleasure, Steven. Anytime. 

Kilger: And for everyone out there listening, thank you so much for doing so. And until next time, stay safe.