When I started building the Transfix platform in 2013, artificial intelligence in trucking was a relatively one-dimensional concept. The Washington Post speculated, “It’s not hard to imagine self-driving trucks streamlining operations for companies that depend on supply-chain efficiency,” while the Commercial Carrier Journal wagered that “self-driving planes, trains and trucks [would] lead supply chain redesign.”

While both might still be right, the focus solely on autonomous vehicles themselves actually sells the more immediate impact of AI in trucking a bit short. Trucking is a $700 billion dollar industry that moves 71 percent of America’s freight by weight. When you look under the hood, there are numerous applications for AI to improve, enhance and expedite processes.

That’s what we set out to achieve in 2013. We’ve made significant strides in laying the groundwork for a more efficient trucking ecosystem. Today, AI and machine learning are being used to fix some of the industry’s most intransigent problems like pricing and route customization, while also making vast data sets useful as predictors of macroeconomic trends. Now, on the cusp of 2019, we can truly see the impact of AI on both shippers and carriers as all sides of the equation become more accustomed to (and reliant upon) 24/7 visibility, robust analytics and a more actionable supply chain.

With that as backdrop, here are three ways that AI will benefit trucking in 2019:


Solving the Complicated Pricing Equation

Factors that drive shipping costs can be unbelievably volatile and extremely complicated, so AI’s ability to adapt to changing circumstances is key. It’s helpful to consider the basic math involved: There are about 150 key markets that serve as trucking destinations in the U.S. Since a combination of any two of those markets can serve as a trucking route, this means that there are, conservatively, about 11,325 trucking routes that are run regularly in the U.S.

The cost of shipping along any one of these routes is influenced by much more than simply distance. To begin with, there is the cost of fuel, itself a volatile commodity subject to wild swings. Then there is weather, the labor market for drivers, and the shifting economics of seasonality. By seasonality, I mean the constantly changing patterns of supply and demand for commodities across the country. The beginning of the orange harvest in Florida, occurring around April, for example, can drive up shipping costs from that region fourfold in a matter of days, and have a lasting impact around the southeast through June. A suddenly hot construction market in Arizona might do the same for that route. Plus, larger economic factors like interest rates, GDP growth and tariffs inevitably play their part in influencing price fluctuations.

Imagine that you run a company whose business depends on getting consumer goods, like clothing, beer, or shampoo, into stores around the country. The health of your business depends on the predictable, efficient delivery of products from, say, 20 manufacturing sites and distribution centers to thousands of retail locations. Each route has shipping prices that constantly change according to multiple factors, so the ability to book fair freight prices, quickly, to guarantee the on-time movement of your products is essential.

The sad reality is that most trucking rates today are still decided in the same way that they were 70 years ago—by people slogging through hours of phone calls to finally get two parties to agree on a price. With AI, however, we are already seeing vast improvements compared to current techniques. Artificial intelligence is able to automatically and efficiently determine fair, reliable prices for thousands of shipments in less than a minute. These processes are currently enhanced by what’s called a “human-in-the-loop system,” in which a human expert helps train, adjust, and validate prices in the presence of events that are unfamiliar to the AI model. These events might include natural disasters, newly-arrived cargo ships, or even geographic peculiarities like steep inclines on a given route. The AI system and human expert work together to generate fair prices that adjust with the constantly changing freight market.

The result is that a process that previously took untold hours of manual labor can now be done automatically and at-scale in a matter of minutes, saving massive amounts of time and money.


Perfecting Preferences

Carriers are the companies that actually own trucks and make their living by delivering goods from one place to another. Truckers, of course, are the folks that drive these trucks. To make their livings, both groups suffer through rigid, archaic processes. On any given day, truckers and carriers field dozens of phone calls or wade through Craigslist-like load boards to find their next shipment, often deciding on the fly if that shipment is going to make them the most money or prove the most convenient for their time.

AI can take this burden away. By learning over time which routes are the most profitable, save the most time, and put truckers in a market ripe for a profitable return trip, AI gives carriers the ability to make more money with less work.

AI can even learn truckers’ individual preferences over time, in the same way that Netflix learns to recommend new shows. Some truckers might prefer a less profitable route with more polished rest stations along the way, while others may want to stay away from congested city areas. AI can make sure that they get served only the routes they want.


Serving as a Barometer of Economic Health

Since trucking is so integrated into the larger economy, freight trends serve as leading indicators for the health of dozens of industries. For this reason, AI-driven improvements inevitably lead to better economic data at both the macro and micro levels. Manufacturers, grocers, financial analysts, economists, and anyone else who can benefit from the insights provided by more comprehensive, accurate shipping data, will find their jobs made easier.

For example, when the demand for trucks in Los Angeles heats up, this often means that imports from China have arrived and trade has increased. Freight data can help predict demand for the heartland’s corn crop, Detroit’s automobiles, or the Pacific Northwest’s lumber. As this data flows through the economy, decision-makers at every level will make smarter, more informed choices.

So, make no mistake: Those of us who are lucky enough to work in the trucking and logistics industry are making full use of AI technologies to improve how fast, and how well, we do our jobs. Pick-and-pack robots, autonomous trucks, warehouse drones and other more sci-fi and eye-catching technologies are being rolled out lightly and tested in all areas of the supply chain. Perhaps the more mundane applications like instantaneous pricing and route customization are not as well known, but they are the meaningful ways in which AI is presently affecting fundamental and much-needed change.

– Jonathan Salama, Co-founder and CTO of Transfix

We stand in solidarity with the Black community. We condemn all acts of racism, discrimination, and senseless violence.Learn More