TMS Suppliers Prepare for AI Implementation
[Stay on top of transportation news: .]
As tech startups roll out virtual agents and assistants powered by artificial intelligence, suppliers of transportation management systems are building connections with those new capabilities while also investing in their own AI features.
“The possibilities are both endless and exciting,” said Ben Wiesen, president of Carrier Logistics Inc., a TMS provider for less-than-truckload operations.
While AI is already an important part of the applications that the company provides today, this technology is advancing rapidly and the industry has only just begun to leverage it, he said.
Wiesen said his company sees “tremendous opportunity” for AI agents to improve functions such as dispatching, customer service, data entry and accounts receivable and payable.
¾
“The opportunity for efficiency gains and for cost savings are so significant that companies who don’t have an active plan to start integrating AI agents into their workflows are going to be at a competitive disadvantage,” Wiesen said.
The implementation of AI agents also could bring significant changes to the jobs and job functions within the trucking and logistics industry, said Jonah McIntire, chief product and technology officer at Trimble.
“The AI agent movement is a completely different class of technology from everything we had before, because it is really competing with labor,” he said. “It’s not competing with other software.”
AI agents currently have both advantages and limitations or drawbacks compared with manual labor.
“They’ll work 24/7,” McIntire said. “They don’t require a lot of money. They don’t have to be paid very much, but they’re kind of savant-like. They’ll make strange mistakes right now, so you have to [maintain] guard rails.”
Trimble’s strategy within its transportation segment is to start by transforming its own business by working to automate certain roles.
“If we succeed, there’ll be no more humans in that role,” he said.
People currently working in roles that become automated may shift to another position and take on different work assignments. In fact, their careers may even accelerate, McIntire said.
The company’s goal at the start of this year was to automate between 4% and 17% of the roles held by human workers.
“We are above the 4% already,” McIntire said.
ѳԳپ
Successful AI implementation in the transportation industry will hinge on integrations and access to the right data, according to Hans Galland, CEO of cloud-based TMS provider Beyond Trucks.
“AI can only be as good and as fast as the data that’s available,” he said.
To improve business operations, AI programs will need to be running in front of users at the moment when they make a decision, Galland added.
He likened it to serving a consumer, when purchasing an item online, being given the option to spread payment over four months.
“Only because you were presented with that option in that split second [can the shopper] exercise that option,” Galland said.
The Beyond Trucks TMS is built with one program, or base code, that interacts with segregated databases.
More Q2 iTECH
►How AI Could Help Streamline Logistics Operations
►TMS Suppliers Prepare for AI Implementation
►Tech Vendors Roll Out AI Features at TIA
►Dysart:How Trucking Companies Can Harness YouTube Tools
“The advantage of that is you get one big program that constantly can evolve and improve and update,” similar to the way phones are updated on a regular basis, Galland said.
Through that method, he said, “you make smaller improvements to the software, but more frequent improvements. So over time you actually develop much faster.”
Having one program makes the building of integrations with other programs very efficient, Galland added.
Along with integrations, TMS vendors also are developing their own AI products.
McLeod Software is preparing to roll out its first AI-driven application, MPact.RespondAI, which reads, classifies and prioritizes communications from email inboxes and telematics systems.
By integrating with McLeod TMS data, RespondAI automatically drafts responses, which accelerates response times, the company said.
“We had all the data from a driver communications standpoint,” said Doug Schrier, vice president of growth and special projects at McLeod. “We saw that the average response time for an open-ended question to their operations team was taking over 40 minutes.”
These questions include requests for a reference number to deliver a load, asking for final directions into a logistics yard, updating an estimated time of arrival or moving a scheduled appointment.
Want more news? Listen to today's daily briefing above or go here for more info
Carriers and brokers want to respond quickly to these types of inquiries, “but freight is much like a thousand gnats flying in your face,” Schrier said.
Developing the new AI application to shorten response times involved designing it to ingest TMS data fed to it automatically, interpret the questions, generate responses and allow a human in the loop to validate the responses.
It can take six months for a carrier or broker to bring a new worker on board and two or three times as long for that worker to become proficient in every activity associated with the job, Schrier observed.
“We think we can help shorten that time by providing more insights and helping trigger that next most important thing,” he said.
