What's really happening with artificial intelligence in transportation, warehousing and logistics? Jim Lee, principal with the ZS consultancy, helps us to separate hype from reality.

Transportation, warehousing and logistics companies today are expressing multiple concerns that could potentially be addressed by AI, Lee says. Top ones include labor and transportation volatility, along with siloed execution as an “honorable mention.” “That hasn’t been effectively addressed,” he says.

How AI should be applied depends on the type of work involved. When it comes to labor, the technology can take over certain “menial” and repetitive tasks, but other responsibilities are better left to humans, Lee says.

Warehouse labor accounts for between 50% and 70% of variable expense that drives cost per unit in the distribution center, he says, so there’s “a ripe opportunity to get some real value” from AI. Applications include optimizing the scheduling of outbound shifts and allocating resources to best slot products and pallets. “It all comes down to increasing the productivity of the travel path and replenishment time,” Lee says.

On the transportation side, AI is ideal for aiding in the highly complex task of route planning and optimization. But exactly where AI would be of best value in logistics varies widely among settings. No two warehouses are exactly alike, and the type of AI to be deployed must also be determined.

For logistics operations unfamiliar with AI in any form, it can be intimidating to approach the technology for the first time. “First and forecast,” Lee says, “it’s the return on AI investment.” (Hence the acronym “ROAI.”) Users must be sure that they have established a business case for the use of AI in every instance. Failing to do so until a project is halfway along is a recipe for failure. And such calculations must be made at every step of the way. “Really it’s just a muscle that we have to understand more,” he says.