AI provides critical assistance to TMS companies

[Stay on top of transportation news: Get TTNews in your inbox.]

Technology vendors say the inclusion of artificial intelligence in transportation management systems (TMS) is changing the way systems work and how trucking companies and their partners use them to increase efficiency and reduce costs. It is said that They also argue that AI also provides predictive capabilities to help carriers and partners identify potential revenue streams and opportunities.

Mercurygate co-founder and chief innovation officer Steve Blow said one of the most compelling benefits is the reduction in air miles in the supply chain. “AI allows shippers and their partners to stop the movement of ‘air’ on highways,” he said, which also helps reduce carbon emissions. “We use AI to determine which shipper’s cargo best fits into the trucking company’s preferred lane, and minimize empty movement by determining alternate shippers to line up for empty stretches. increase.”

You can train AI and machine learning models to support decision making and automate processes. Jonathan May, director of business intelligence at McLeod Software, said McLeod Software is working on a model that will enhance its capabilities in fare evaluation, carrier payments, EDI bid acceptance, planning and scheduling, and driver matching. said. Initial investments in AI (hardware, software and labor) “can be expensive, but the efficiency gains and cost savings will be his ROI,” May said. A TMS vendor is incorporating chatbots built on his OpenAI API from Microsoft Virtual Agents into its pricing, lane and rate analytics products.

McLeod Software has integrated truck load optimization solutions from Manhattan Associates and Optimal Dynamics into its TMS, said Robert Brothers, vice president of product development. Brothers said users will be able to take advantage of optimization recommendations from these products in a “regular, role-based workflow without exiting our system,” adding that “the day when real-time traffic input will certainly come to fruition. is approaching,” he added.

Tai Software has built an AI mechanism into TMS for managing emails. Read and reply to emails, link to shipments if relevant, notify staff and generate responses that can be put on hold until someone approves. for shipping. Walter Mitchell, his CEO at Tai Software, says that using natural language processing, TMS can create shipments “from what users submit.” Natural language processing is a branch of AI that enables computers to understand text and speech.

Mitchell said a shipper could send an email saying they want to create a package from “my warehouse.” A language processing tool identifies the customer based on the source of the email and the phrase “my warehouse”. “We have information about them, so we can use AI to predict what the package will look like,” Mitchell said. Tai’s system creates a shipment, generates a quote, and notifies staff when the quote is ready for review. “We literally get a quote back in minutes,” he explained.

Most TMS providers agree that the introduction of ChatGPT by AI company OpenAI has enabled the creation of AI and machine learning products. ChatGPT is known as a “large scale language model” and is designed to predict the next word in a sequence, enabling automated communication between systems.

Tai Software has integrated Parade, an AI-powered freight capacity management platform, into its TMS. Tai also integrated greenscreens.ai, another tool with AI capabilities. Mitchell said greenscreen.ai helps generate pricing. According to Mitchell, this will allow the company to leverage pricing data within TMS when booking and managing shipments.

Concept Logistics installed Tai Software’s systems more than three years ago and is leveraging their integrated AI tools, said Greg Finnerty, Concept’s vice president of operations. Finnerty said the logistics company has expanded its carrier network in the meantime. “Thanks to Parade and Tai, we are able to recognize where we are leveraging our carriers more effectively,” said Finnerty, who said that this allows Concept to improve, for example, load consistency. can be improved.

Chris Orban, Trimble’s vice president of data science, emphasized the importance of defining AI as “computer algorithms that use large amounts of data.”

“I think it’s very appropriate for our industry. There is so much human component that it’s hard to let computers do all the ‘work’,” he said. “Decision-making still needs a little bit of the human element.”

Orban added that AI is very powerful, and computers can indeed examine much larger amounts of data in much less time than the human brain. He said Trimble’s technology can incorporate real-time weather and traffic conditions from external sources to calculate accurate routes and arrival times.

Orban warned that one of the drawbacks is that “computers are very bad at predicting and learning things that haven’t happened before.” He cited the COVID-19 pandemic and the accompanying supply chain disruptions. As states closed rest areas for public health reasons, truck drivers had no place to stop and rest. In such cases, predictive solution algorithms are useless. “You can’t train a computer to solve a problem it’s never seen before,” Orban said. “It’s an important limitation.”

McLeod dispatcher. McLeod can alert customers to certain data integrity issues and set default values ​​when data is missing. (MacLeod Software)

Another challenge, technology vendors say, is ensuring clean and accurate data.

“Data integrity starts with the user,” May said. “You can curate and cleanse the data that you use in your model. You can also impute missing values ​​based on similarity scores or other logic,” McLeod said, adding that the data can impede decision-making. It warns customers about certain data integrity issues that it may pose, adding that it sets default values ​​for missing data elements. “We are also looking at adding more drop-down, menu-driven data points within the system rather than free-form fields,” he said. “It’s useful to use standardized data tables instead of free-form data fields. Free-form data fields can introduce a lot of ‘noise’ and variation in your data. May said it’s important to present clear information in context and at the right point in the user’s workflow.

With regard to “system-generated” output, trust is also an issue, he said. “[Customers ask]”Should we believe this number, or should we follow this suggestion that the system presents?” May said. “To build trust, it’s important to provide feedback on how the models and the advice they gave actually worked in real-world scenarios,” he said.

Many TMS providers use DDC FPO technology to ensure clean and accurate data.

DDC FPO specializes in business process management, using agency teams or AI solutions, for example, to process data on freight rates for carriers, said Madison Conway, global marketing director at DDC Group. . “Often we are integrated with TMS,” she said, using it by shipping and logistics companies. “We have tens of thousands of bill of lading templates in our database,” she said.

The company’s AI-based data capture software, DDC Intelligence, matches incoming BOLs to templates in its database or learns and saves them as new templates. The next time such a template arrives, “it’ll be completely hands-free and direct,” Conway says. The result, she said, is more accurate billing data that can be submitted to shippers “for faster payment.”

Ben Wiesen, president of Carrier Logistics, which provides transportation management software for LTL carriers, said the industry is turning to AI and machine learning to address routing, sequencing and even staffing issues. . “The computer helps a lot, so maybe he’s one person and he can handle three terminals. [by taking on more decision-making],” He said.

Wiesen said the computer systems in the LTL field have, for the most part, “done a great job in the initial planning,” optimizing the pick-up and the best way to do it. “What computers don’t do well is reacting to changing conditions on the ground,” he says.

CLI is working on applying AI to look at every truck in the LTL fleet and suggest ways to save, say, 6 miles by reassigning one pickup to another route.

“It takes a lot of data analysis to make that decision,” Wiesen said, as it has to monitor all trucks, pending pickups and conditions on the ground.

Need more news? Listen to today’s daily briefing below or visit here for more information.