There are many “buzzworthy” topics in the transportation and logistics sector that are often discussed, but less frequently showcased in action. Artificial Intelligence (AI) and machine learning (ML) are two such phrases, each of which were the topic of a recent session in Trimble’s Supply Chain in.sights Virtual Series + Interactive Community Event.
Featuring panelists from Covenant Transport, Inc., and Paper Transport, Inc., Trimble leaders hosted a presentation called “Beyond Buzzwords: Applying AI and Machine Learning to Transportation,” as part of the “Elevate Your Data Perspective” series.
At their core, AI and ML are simply methods for analyzing data from the past to make predictions about the future, says Peter Covach, director, industry solutions advisor, for Trimble Transportation. Using advanced mathematics and historical data, data scientists can develop models for making predictions with a high degree of certainty.
Throughout the session, Matt Mullins, vice president of program management for Covenant Transport, and David Dunst, IT Manager for Paper Transport, shared examples and learnings of how their trucking fleets have harnessed these technologies to improve efficiencies and enable data-driven decisions across their organizations.
Mullins says Covenant Transport is currently using a Business Intelligence tool called Domo as a stepping stone to more enhanced AI and ML operations across the entire enterprise.
Three major areas where Covenant Transport currently uses AI include:
Resource utilization – analyzing areas of opportunity within specific markets
Warehouse optimization – reviewing trends in which SKUs are most popular and revising warehouse layout accordingly
Robotic process automation (RPA) – automating repetitive tasks across the organization
Driver retention – keeping an eye on data points that can help drive positive outcomes around retaining drivers
Because it is a relatively easy tool to implement, in less than a year since the solution was rolled out, the number of users within Covenant Transport has expanded to over 200. Mullins says the real-time feedback has been very beneficial to staff, assisting them in making more informed decisions. The solution can connect to databases in many areas of the business, so Covenant Transport aims to roll it out more widely across the organization in the future.
Dunst of Paper Transport explained that the company has used a variety of approaches to implementing AI in various areas of the company. For example, Paper Transport built its own “Accident Prevention” predictive model to analyze risky driving behaviors. Dunst said it took a good deal of internal resources to accomplish, but has helped safety managers better understand which drivers are at risk and how to help coach them.
Next, Paper Transport partnered with Trimble to implement Trimble Dispatch Advisor, an optimized dispatch solution that prescriptively suggests matches between loads, drivers and equipment. The Paper Transport team reports that load planners enjoy the intelligent guidance the solution provides, while ultimately still giving the power to make matches and assignments to the planner, who may know something the system does not.
“As powerful and as cool as these terms, AI and machine learning, are, we as humans might know something that the computer doesn’t,” explained Chris Orban, vice president of data science for Trimble. These variables might include driver preferences, or scheduling requests that a dispatcher might have in mind as they’re reviewing recommended loads.
Dunst says the carrier has also made significant strides with its RPA solution, which is helping reduce the amount of menial, repetitive tasks that employees weren’t interested in doing, and allowing them to focus on more meaningful, engaging work.
Trimble’s Approach to AI and ML
In addition to the Trimble Dispatch Advisor for optimized load matching, Trimble offers several solutions that utilize AI and ML.
Trimble’s Vehicle Health solution offers a more proactive approach to maintenance activities. An AI algorithm analyzes thousands of sensor values across the range of Trimble-powered vehicles to identify values that are outliers based on current conditions, such as weather, area of operation, type of operation and more. These variables can greatly affect the performance of a truck, so having advance warning on a potential maintenance issue before a fault even fires can help fleets be even more proactive with maintenance activities to keep trucks on the road.
Automated Driver Coaching is another area where Trimble is applying machine learning. By combining forward-facing video with telematics data from the engine, Trimble can provide a customized, behavior-driven coaching solution to drivers, after each driving task. At the end of the day, the algorithm can select specific coaching videos and recommendations to tailor a safety program for a driver that identifies both good driving behaviors and those that may need improvement, within a time frame that is much closer to a specific incident that may need review and coaching.
As with any AI or ML algorithm, the more data you have, the more accurate predictions and recommendations will be. That’s why Trimble, whose customers generate more than 10 billion data points per day, is able to develop deeply insightful and actionable solutions that are specific to the trucking and transportation industry – allowing algorithms to simplify and optimize menial, repetitive tasks and helping people to focus on the work they do best.
Artificial intelligence and machine learning models continue evolving over time – as do carriers’ needs for analyzing their data. We’ve only begun to scratch the surface for potential uses of these technologies, and many efficiencies and more data-driven decision-making are already on the horizon.
Gain Supply Chain in.sights: Watch a Replay of the Conversation
Looking for more helpful “in.sights” into the important role of these technologies and practical tips on how to apply them to your operations?
Watch an on-demand replay of the full discussion to get an in-depth look at how artificial intelligence and machine learning are optimizing all parts of the transportation supply chain.