“Twenty years ago, we could not build [and operate] software with the capacity to store and access huge caches of historical information and data and calculate [things like] 10-dimensional optimization,” recalls Pawan Joshi, chief strategy officer for
e2open, a leading developer of transportation management software. “We didn’t have the data or the computing resources to build these decision-making models.” With the advent of artificial intelligence and the extremely powerful computing resources behind it, “now we have the computing power with the speed to do it.”
A CONTINUING JOURNEY
Srini Rajagopal, vice president of logistics product strategy for
Oracle, sees AI as just the latest step in the continuing journey of maturity and innovation in the TMS space. He breaks the development of AI into two parts. “The first is the standard, classic AI model. These support specialized [computing and analytics] models built for specific purposes,” such as developing optimization and consolidation plans, routing or ETA predictions for trucking, or cycle-time predictions for warehouses.
The next step is “generative AI, which has come about because of the maturity of the large language models (LLMs) now available,” he explains. This development allows the software to interact with users in a natural language format, creating new opportunities for task automation in the typical cycle of transportation planning, execution, and exception management.
“What we use that for is [to give the model] the ability to interact [with a user] in a natural language format and then do reasoning about what actions to take [based on the user’s input].”
He cites as one example the returns process, where typically a customer service agent will engage with a customer and answer questions over the phone. “The AI agent can take over a lot of that role, responding to the customer’s questions by voice and making recommendations based on the user’s input.” That frees up time for the human agent, who now may have to intervene only with a small portion of questions that the AI agent cannot handle. “Now the human agent has more time to focus on other, more complex or higher value-added tasks,” he notes.
ROI STILL RULES
Yet even with the advent of more advanced and sophisticated machine learning algorithms and artificial intelligence taking on more complex tasks, at the end of the day, “when it comes to execution, that’s where the rubber meets the road,” says Oracle’s Rajagopal about the principal role of a TMS and the realizable and measurable results it can provide.
That should be the priority, he notes: Value measured, quantified, and validated across numerous metrics—whether it’s lower operating costs; more efficient, less error-prone processes; better transportation procurement; or optimized and more productive use of assets and people.
One shipper cites his rule of thumb for ROI (return on investment) as being “for every dollar spent on a TMS annually, it should return at least $2 in direct annual cost savings and/or productivity gains.”
Those gains can be measured in a host of ways, notes Rajagopal. “It might be something as simple as billing accuracy,” he says. “Are you getting paid accurately for your services, billing correctly, eliminating duplicate bills?” Then there are what he calls the “soft” benefits, such as user productivity and time savings from automating tedious, manual tasks. “Is your dispatcher or planner able to do more in a day with the new system?” he asks.
“ROI is all about knowing how you were doing before, quantifying the as-is state and what it costs you, and then, as you implement, measuring what it looks like in the new state and validating that you, in fact, got the savings expected.”
CONNECTIVITY AND VISIBILITY
Tom McLeod, president and chief executive officer of
McLeod Software, has spent decades helping truckers and brokers use technology to work better, smarter, and more efficiently. Over those decades, he says, two demands from customers have remained constant: connectivity and visibility. “That’s been an ongoing theme in technology development for our industry in the last 10 years,” he notes.
He sees AI as a tool that will streamline the exchange of information between shippers and carriers, ultimately improving the executional accuracy and efficiency of the transportation planning and execution lifecycle.
One key foundational aspect of achieving that goal is integration and how effectively and seamlessly companies like McLeod and other TMS operators can help customers accomplish and maintain that. It’s a continuing challenge that gets more complex but also is benefiting from technology advancements that make the task both simpler and faster to accomplish.
“We have seen a real explosion of integration requests and requirements,” McLeod says. “More and more companies are coming into the market providing information services, and the pace of change is accelerating.”
McLeod’s focus has been “to offer the … best integration to our customers so that they have a chance to compete. And to have an open platform that enables them to do so,” he says, adding that “once it’s complete, that process needs to be automated, with the information going where it’s needed, and being accurate and reliable.” And for the technology providers to be adaptable as the industry continues to change and new solutions come on the market.
McLeod supports this strategic imperative through its Certified Integration Partner program, which offers off-the-shelf, supported integration solutions for over 180 different trucking industry software products or services, from over 130 different companies.
Even with the advances in TMS platforms, in the trucking world, there are still “a lot of niche markets that require almost totally different services” as well as a lot of repetitive, manual tasks still waiting for automated solutions, says McLeod. He sees significant opportunities for TMS providers to help customers truly re-engineer their operations, addressing important metrics such as reducing deadhead miles, increasing revenue per mile, and getting more revenue per employee.
“It’s not for the faint of heart,” he adds. “As apps get more sophisticated, it is important for us to continue to handle more and more details, on a more automated basis. That’s what carriers want and need to help them better serve their customers, keep costs in line, and compete.”
Nevertheless, with all the promise of technology and the opportunities for AI to accelerate the shift to automation, “it is still a relationship business, between people who need to ship goods and those who provide the assets, resources, and expertise to do that,” McLeod stresses.
“Even as routine transactions are automated, when it is crunch time and there is a problem, people still want to have someone on the other end they can reach out to, that they know and trust,” he says. “Technology cannot get in the way of strengthening those relationships—or replacing them. It must support and facilitate that.”
NO PATCHWORK QUILT
As the nation’s largest broker and freight forwarder,
C.H. Robinson (CHR) has a view of the market—and the role of technology in it—that could certainly be considered informed. With integrated management services that touch every mode of transportation, both nationally and globally, the company has a deep view into the needs and wants of shippers worldwide—and how technology can address those needs.
One recurring theme among CHR’s customers, says Jordan Kass, CHR’s president of managed solutions, is “shippers are not looking for a point solution anymore. They don’t like the idea of a patchwork quilt. They want one pane of glass [through which] they can see and control their entire supply chain,” he notes, adding that over 50% of CHR’s revenues come from customers who use both its forwarding and surface transportation management capabilities, across modes.
He believes that is a function of shippers who are stressed to the max, are coping with a shortage of supply chain talent, “and are being asked to do much more with much less.”
For CHR, he cites as a key advantage its proprietary TMS—which is both global and multimodal—and an engineering team that continually works to improve and expand its capabilities. He also believes the advent of AI will be incredibly transformative for the industry.
“Because we are building [the TMS] and using it at the same time, we have a really unique and valuable eye into how it performs and what customers want and need. As we operate the platform, we identify use cases with our customers and then go to our engineering team to build a solution,” he notes.
Kass says CHR’s technology approach as a builder and operator of its TMS gives it a unique look into “how transformative AI can be in this space and how we can lean into some of the larger problems that shippers are dealing with.”
As one example, he cites CHR’s development and implementation of “touchless” appointments for freight pickup and delivery. “If you think back, making a [pickup or delivery] appointment used to take multiple tries [with phone calls, texts, and emails], and it sometimes required more than a day to get that appointment in place,” he recalls.
With its AI-driven process, “now we are doing that in under two seconds, greatly enhancing the speed of that process and adding huge value to it.”
CHR has data on 35 million shipments a year, Kass says. That data informs the AI engine, which in determining the ideal appointment time, will consider things like patterns in transit time along a route, on-time performance, and dwell time at a facility. It will even take into account what’s ideal for the carrier.
For example, Kass says “carriers in South Dakota need a longer time to get to the point of origin because they’re typically traveling farther, so a 6 a.m. pickup appointment isn’t good for them, while a 6 a.m. pickup appointment in an urban area might be great for a carrier because it can avoid traffic. The data [accounts for] these things better than a human can.”
One area that TMS providers need to improve upon is predictive capabilities, Kass believes. With AI, “as you feed more data into the system, the more accurate you get.” With that come more opportunities to expand the platform to automate and streamline tasks that continue to be done manually. It also helps the TMS get better at interacting in real time with transportation processes and accurately predicting outcomes. “We have the scale, and with AI, the more you feed it, the more intelligent it becomes.”
IT STILL COMES DOWN TO COST
Even with the inexorable march of technology, its permutations of AI, and its promise for positive change and automation that helps its human partners work smarter, faster, and better, in the end, it still comes down to cost—measuring and weighing what’s being spent on the TMS against the operational cost savings and productivity being realized.
“The shipper’s main concern is still cost,” says Bart De Muynck, principal at consulting firm
Bart De Muynck LLC. “That comes from a couple of areas. One is to better optimize the freight spend. Second [is to] put in a better process for the shipper to tender freight to the carrier and for the carrier to [handle] that freight in the lowest-cost manner possible. [Yet another is to obtain] transparency, providing better insights into how the shipper is procuring capacity so shippers end up with reliable, quality capacity at the most affordable rates.”
And as technology has become simpler to integrate, implement, and use, “everyone can and should buy a TMS,” De Muynck says. “There are many flavors; they have become more intuitive, faster, and easier to use.” It’s not about offering completely different things, he adds. “It’s about streamlining the user activity and how the systems perform everyday tasks, making the job easier, and making it easier, more convenient, and less costly for the shipper to work with the carrier.”
Not so fast …
After seeing the possibilities of what a TMS can do, companies sometimes will be in a rush to get their solution implemented and operating. That can be a mistake that leads to errors and an unsatisfactory outcome, says Keith Whalen, corporate vice president of product management for TMS provider
Blue Yonder.
Shippers should make sure they take the time to “focus not only on the really important cost savings, but also, when you scale volume, on doing performance testing” to ensure assumptions are holding up and performance meets expectations, he notes. “Not just [testing] the initial design and integration, but having a more holistic view in all areas, leaving adequate time and not rushing through. Don’t skip steps,” he advises.
Whalen counsels customers to spend the time and effort up front on knowing their current state, modeling out what they want the future state to look like, and, importantly, planning for training and change management to bring users who will be operating the platform successfully into the new realm.
“I think one of the things we do a really good job at is up front in the initial modeling,” he notes. “The customer should be examining opportunities across its transportation network [and] do ‘what if’ analyses to look not only for savings, but also at where [it might get] the biggest bang for the buck.” Such efforts might look at a nearshoring strategy and how it changes the supply chain, a decision on fleet asset deployment or type of service, or warehousing locations to optimize the network and respond to a shifting supply chain.
“That modeling and initial ROI calculation builds the business case. It not only justifies the deployment of the TMS, but also provides the guidance on how to roll it out as they go through their projects,” he notes.
Lastly, he stresses that training the operating team, helping them change and evolve from past practice, and transition effectively to the new tools, can be the difference between success and failure.
