March 17, 2018

PRODUCE TRANSPORTATION: Data-Driven Predictive Models Could Aid Industry Down the Road

Using information provided by many different shippers and a computer program, the Western Growers/C.H. Robinson Transportation program has brought increased efficiency, better service and cheaper rates to shippers enrolled in its LTL (less than full load) program.

Aggregating these LTL shipments has allowed the program to schedule transportation on a reoccurring basis, allowing a shipper with even a partial load to take advantage of completive rates. Heretofore, a load requiring multiple pickups and multiple drops was the scourge of the industry. No trucker wanted to participate without premium rates. And the new electronic logging device regulation makes the proposition an even harder sell. But access to data has allowed CHR schedulers to note that several shippers have LTLs headed to Chicago, for example, on a regular basis. Loads can be scheduled and efficiencies can be achieved. Collaboration and cooperation is the key…and it works.

But still Luke Gowdy, general manager of operations for C.H. Robinson Worldwide, said this type of advanced planning is only occurring three to 10 days out. Imagine a future, he said, in which data science and artificial intelligence can be used to always make sure the capacity is where it is needed, to be able to accurately predict, with a great amount of certainty, how long it will take a load to get from point A to B and where it might be most advantageous to pick up a shipment if there is a choice. Might it be better to pick up a load in Santa Maria rather than Salinas because of an unrelated event snarling traffic?

Charlie Loes, director of fresh technology for CHR, said the company is already moving in that direction with the introduction of its Navisphere Vision platform last fall. The effort is on the front-end of creating a predictive model for the transportation of fresh produce and other items. Loes though is candid in noting that the accuracy of such predictions improve as the number of data points increase. “Our logistics operations has generated three billion unique data points over the last six years, and the Western Growers partnership has made a significant contribution to that,” he said.

While that is seemingly a great deal, it’s only the tip of the iceberg. In a discussion with Western Grower & Shipper on the future of fresh produce transportation, Loes and Gowdy said transparency is the key for the most efficient use of transportation. If everyone had access to all the data surrounding the shipments of fruits and vegetables from California to the rest of the country, a very efficient transportation system could be managed. The use of a computer and artificial intelligence could much more efficiently schedule those 500 loads that might be leaving the state with lettuce for the Midwest today.

The produce industry, rightly so, prides itself on its unpredictability. Every day is different is the mantra heard from everyone along the supply chain. Loes said that while that is true on the granular level, zoom out and nine out of 10 loads are reoccurring and predictable. If only all the data was available to a transportation scheduler, a better system could ensue. Bottlenecks could be identified and avoided. Pickups and deliveries could be better scheduled to decrease wait times at each end.

Gowdy said there will always be a need for spot market brokering of a truck. A shipper will make an unusual sale to a new customer in a different place and will need a truck quickly. But that is the exception.

“We just want to move the industry forward to a more predictive model,” he said.

Of course that is already occurring for many shippers participating with CHR and other logistics companies to fill their contracted sales. Many companies have standing orders from buyers, which allow for the advanced scheduling of equipment. But Loes said much more could be done. More data, he reiterated, leads to better analytics and allows transportation service providers to better cater to the needs of their customers.

Of course, these two industry veterans understand that complete transparency might be too high of a hurdle to jump in the fresh produce industry. Industry members believe they are uniquely independent and are really not that interested in sharing all of their data. Gowdy said CHR knows this will be an evolutionary process, which has already begun. He said the history of the WG/CHR program shows a great deal of technological advances over the last decade. “We are absolutely doing a much better job,” he said, noting that the data accumulated does allow for better positioning of equipment. He added that on a year-to-year basis trends are being analyzed and acted upon.

But Loes said data science is changing industry after industry and there is no reason it wouldn’t do the same to the transportation of fresh produce. Ten days, 30 days, six months out, the industry should be able to better manage its transportation equipment needs, and shippers should be able to make smarter choices. He said Navisphere Vision is already performing some predictive tasks—like predicating the best route for a cross country trip taking into account, weather, traffic and other factors—but he has no doubt it will do more in the future.

In the short term, CH Robinson is still adept at using some tried and true methods to forecast the needs of its clients. Gowdy said looking forward to the normal rise in fresh produce shipments in the spring and summer this year, it’s clear that there will almost certainly be some volatility. “There are macro indicators that supply and demand (of equipment) are not in balance. We are continuing to see an increase in demand, and for many reasons we see less supply of equipment.”

Currently, he said truck utilization is as close to 100 percent as you can get. So as there are surges in supply, there is a high level of certainty that there will be periods of shortages and high prices, especially on the spot market. Add in the regulatory components, including the electronic logging device regulation that will, in aggregate, decrease the number of trips for the average driver, and shortages can be predicted. He said the transition in supplies from the desert production areas of Arizona and California to the valleys of California will certainly result in surge volumes and probably surge pricing.

Looking at the upcoming situation through the lens of what occurred when increased equipment demand took over after the hurricane, it is almost inevitable that shortages will occur. But Gowdy said CHR did learn from that event and is doing what it can to position carriers as effectively as possible. He said the best strategy for shippers is to align upcoming freight opportunities with capacity, which basically means committing shipments to specific carriers.

He also commented on the many reports that truck orders are up significantly year over year leading to much speculation that supplies are catching up with demand. Gowdy said in the first place it takes a while for a truck order to get into commerce. These orders are placed by truck dealerships but the truck still has to be sold to a company or an individual and then get on the road.  “We know there is a shortage of drivers. Just because there are trucks purchased, doesn’t mean the driver supply is there.”

Ultimately, it could be a challenging spring and summer. It points again to the need for better predictive models so those challenges can be anticipated and met.