Inbound Logistics | July 2023

Validate the information. Advancements are quickly rolling out to improve large language models (LLMs), but users should be cautious of using these systems for important decisions without domain expertise validation. –Alex Schwarm, Head of Data, Arrive Logistics

“For example, Copilot (editor’s note: not to be confused with Loadsmart’s CoPilot) in the Microsoft Supply Chain Center can identify poten- tial supply chain disruptions like natural disasters, civil unrest or geo- political clashes, or lagging economic indicators. From there, Copilot can automatically ag these events, show the users which customers are impacted, and create email messages to reroute purchase orders or check-in on customer contacts,” notes Sigler.

for buyers of marine equipment, services, and solutions. It recommends suppliers through its machine learning algorithm and marketplace of service providers to make the purchasing process faster and more streamlined. “When managed properly, arti- cial intelligence can improve supplier relationships through real-time data vis- ibility,” says Spencer Shute, principal consultant at Proxima. “AI can allow companies and their suppliers to work in harmony, reducing the risk of supply and demand inaccuracies and offering a transparency that is difcult to achieve with competing priorities.”

storage space and labor availability across their sites. For instance, the LevelLoad solution from ProvisionAI analyzes shipment pat- terns and identies spikes in demand over the next 30 days. The system can then adjust by shipping some products early or holding less-needed items a day or two. One LevelLoad customer is Kimberly- Clark, where the solution is run nightly, says Dr. Jeffrey Schutt, chief scien- tist at ProvisionAI. The solution helps the consumer products giant even out loads and “has resulted in substantially higher tender acceptance by their pre- ferred carriers, signicantly reduced transportation costs, and more efcient operations at their distribution facilities,” Dr. Schutt says. AI-enabled short-range transporta- tion planning helps optimize loads and increase efciencies. “Using reinforce- ment learning, we can construct realistic candidate loads very quickly that know that bricks can’t be stacked on eggs and have excellent weight/cube mixes of products to ‘max out’ the loads,” says Dr. Schutt. The solution then creates the required signal for the TMS (transpor- tation management system) to perform early tendering.

IMPROVING RESPONSES TO CARGO THEFT.

Cargo theft prevention efforts will get a major boost from AI, which can help improve incident response times. For example, Overhaul, a supply chain visibility, risk, compliance, and insurance solution, launched an AI fea- ture called RiskGPT in its platform to allow users to quickly respond to in-tran- sit shipment risk. Because RiskGPT’s model is trained on Overhaul’s data, shippers can get an answer with details and contextual accuracy when they ask RiskGPT how to respond to a specic event. They can manage and mitigate a greater volume of risks with the tool. Says one Overhaul client, Kok Ee Lian, global supply chain direc- tor at Dyson: “RiskGPT could be a game-changer with the potential to sig- nicantly improve our ability to manage risk with real-time response and correc- tive action—enabling us to better ensure the safety and security of our cargo.”

STREAMLINING INBOUND LOGISTICS. AI can help logistics ser- vice providers (LSPs)

optimize assets for the movement of shippers’ materials or components from suppliers or vendors to their facilities. For example, a major auto manufac- turer is piloting nuVizz’s RoboDispatch Solution in its inbound logistics opera- tions. In this pilot program, RoboDispatch automates the dispatch process for the movement of full and empty trailers from parts supplier locations to its manufactur- ing plants. Leveraging AI and machine learn- ing, the RoboDispatch Solution matches delivery assets to delivery demand in real time. The solution provides a portal for LSP customers to enter orders. It then automatically assigns customer orders to the most suitable contract or eet driver to pick up and deliver the order on time.

SUPPORTING AUTOMATION.

AI-enabled warehouse robotics technology can help warehouse operators increase the efciency of picking operations. “AI will help make voice picking tech- nology more possible and prevalent. Through a headset microphone, a user can receive instant audio updates on inventory levels, supplier constraints, and order status,” says Sigler. For example, Symbotic, a provider of AI-enabled robotics technology

SELECTING AND MANAGING SUPPLIERS. The request-for-proposal process can be stream-

lined with AI. For instance, Verusen’s Trusted Supply application can improve the match rate for materials requested by manufacturers and other companies, enabling suppliers to respond promptly through its natural-language interface. Another solution that harnesses AI is Procureship, an e-procurement platform

LOAD PLANNING. AI can help companies plan loads and create a more balanced trans-

portation plan so they can work with preferred carriers and ensure adequate

144 Inbound Logistics • July 2023