Artificial intelligence applied to transportation

08/04/2022 | 10 min

A dominant theme of the coming two decades is the arrival of artificial intelligence (AI) into the workplace, including in the data intensive area of logistics. Unlike earlier trends that were enabling staff to do more, AI has the potential and the complexity that it raises two major questions for logistics leaders: Where will we find the talent to deploy AI, and to what degree will AI’s ability to handle activities that we pay workers to do revise our ideal workforce? The rest of the article provides one avenue to assessing where to place AI within a logistics operation, using procurement as an example, and how to estimate the change in roles it offers to the staff around it.

Which areas in logistics and which questions can AI help you with?

AI differentiates itself from automation in that it's not about doing a process faster and with less labor involvement, it's about doing it in a way where there is decision autonomy for the software. So, we need to focus on identifying decision points, and apply AI to these steps. In transport procurement, the sourcing decision ("Who do I work with and what do I intend to buy?") and the transacting decision ("How do I buy it, how do I contract with them, and for how much?") are the main decision points.

Often, after those procurement decisions are made, there’s another decision round on the transportation assignment level that evaluates whether the previously set decisions should apply to all cases, by enforcing transportation contracts with carriers ("How do I assign work within the set contracts? To which transport providers should I assign each load?"), or, if shippers should deviate from them and go to the spot market. If the latter is decided, then we are faced again with some of the same questions: “Who do I work with? How do I buy it? Where is my pricing limit, where to begin and to stop?"

How deeply do you want AI to be applied to your decision-making in transportation?

According to academic research, there are 9 levels at which AI can be applied to decision-making. It can vary from having the application take no actions at all by itself and having a user initiate all actions, to having the application decide everything, making decisions (even the ones you are not aware of), and taking actions autonomously.

This is a scoring criterion for procurement and logistics managers to understand where their companies are at and also where they want to be in terms of using AI for decision-making in transportation.

A key takeaway here is that the state-of-the-art in AI right now is that it can finish your sentences but it doesn’t start them. For example, in a messaging application, when you open it to write to someone, AI doesn’t start typing for you or decide whom you are writing to, but once you’ve started typing then AI kicks in and becomes really good at completing your sentences based on data, patterns, and machine learning. That’s the state-of-the-art of a lot of AI applications right now, it cannot start the tasks for you, but once you’ve started it, AI and machine learning can definitely use the data and behavior patterns to mimic your decisions - even your best decisions - and make just the best decisions you’d do at all times, removing human error from the equation.

Why and how to apply AI to transportation?

One of the main motivations to apply AI to your transportation is to leverage talent. Your team is capable of doing much more and delivering more results to the organization if they are assigned to more strategic and high-value-adding tasks. So, you'd want to move high-value talent from an operational, process-driven work routine to a more strategic array of responsibilities with a deeper impact on optimizing your logistics operations, while AI would take care of the more operational decisions and tasks.  

Because AI is right now mainly able to make our best decisions and finish our tasks (rather than starting them), we need to look at the decision suite and evaluate which decisions fall into this category – of finishing sentences rather than starting them.

The negotiation stage of freight procurement is one of the areas of the decision suite that can benefit a lot from applying AI to it. Here, questions like “How do I structure a negotiation?”, “Do I want to halt or continue the negotiation?”, “And if I do continue it, what paths of concessions do I offer?” are covered, and each of these questions leads to decisions that have a constrained decision space, are triggered by previous processes, and can be judged by their success - so their nature is of helping to finish your tasks.

Another major reason for applying AI to transportation logistics is talent reallocation for high-frequency data-driven decisions that have objective right answers. The reason behind it is because you can both free your staff to work on something else, or because they might not consistently take the correct decisions -  and you want to reduce risk of human error by using intelligent systems.

A great example, in this case, is the decision between assigning your transport load to your contracted carriers or going to the spot market (and for which price/rate). This is a high-frequency use-case scenario as some shippers assign hundreds (or thousands) of shipments a day in the spot market, and this is a data-driven decision that has a correct or best answer for each occurrence. Here, AI should be considered as a talent replacement option, because gathering data, analyzing it, and making the best possible decision is  time-consuming and susceptible to human failure.

Often, companies, at least when it comes to their transportation logistics, have a low level of AI applied to their decision-making. This means they often have an application automatically identifying that a decision is needed, and then asking the user to look at it.

Transporeon's proposition with Autonomous Procurement is to go beyond that, by not only informing the best decisions, but also executing them automatically, and then informing users through reports or alerts. No matter whether you are in a capacity-constrained market where you are forced to go to the spot market, or if you buy freight on the spot market strategically, Transporeon Autonomous Procurement matches AI, machine learning, and behavioral science to find capacity at the spot market faster and cheaper (at times, cheaper than your contracted rates), and assign the transport load to the transportation company with the best deal. With Autonomous Procurement as one of the top options in your assignment route, you are set to match up to 90% of your loads automatically, free your talent for more strategic tasks, use spot buying strategically, and save up to 12% on it.

At the end of the day, companies have to decide on which level of AI to apply to their decision-making, but it’s important to know that using it will definitely help improve the efficiency of their freight procurement, assignment operations and, consequently, find capacity easier and make big savings.


Jonah Mcintire is the Director of Procurement Products at Transporeon and draws on over 15 years of experience in the supply chain industry and its digital transformation. Jonah is a big believer in data-driven processes and decision-making, being one of the innovators behind applying artificial intelligence and machine learning to freight procurement and assignment.


AB InBev Brews Up Spot Freight Savings with Transporeon Autonomous Procurement

ABInBev chose to rewrite its own history when faced with new levels of volatility resulting from the global pandemic. After a 45 day pilot proved double-digit savings were possible on spot buying, the multinational drink and brewing company was more than ready to embrace a change. Now, Anheuser-Busch is one of the best spot buyers in the freight market.

Anheuser-Busch (AB InBev)
Largest spot buyer in the world swaps best-in-class proprietary process for Transporeon Autonomous Procurement.
of millions of financial impact.
Double digit
savings in a highly volatile freight market.
reduction in FTE requirement.
direct reduction on spot rates.

Interested in knowing more about Transporeon Autonomous Procurement?

Click here to know more or submit the form below to talk to a specialist and get a demo.


Explore our digital freight solutions

Together our products work in harmony to increase transport efficiency along the full lifecycle of freight activities.

Freight Sourcing Hub

Autonomous Procurement

Autonomous Procurement
  • Automates procurement using data and behavioural science.
  • Analyses how carriers make pricing decisions.
  • Achieves requested capacity at lower freight rates for road transports.
  • Fully automated process of predicting, framing offers, and concluding assignments.
  • Entirely carrier specific and automated process.

Transport Execution Hub

Best Carrier

Best Carrier
  • Access the spot market more easily.
  • Cut transaction costs by up to 19%.
  • React quickly to market fluctuations.
  • Improve process efficiency with better integrations.
  • Cloud-based system provides real-time transparency.

Transport Execution Hub

No-Touch Order

No-Touch Order
  • Automated shipment execution processes.
  • Fewer empty runs.
  • Cut process costs by up to 30%.