Artificial intelligence is the hot technology topic of the moment, especially as it relates to the supply chain. But is it really translating into tangible benefits right now? Sriram Nagaswamy, executive vice president technology with FourKites, offers his view. 

Shippers face a couple of big challenges when seeking to implement AI in supply chains, Nagaswamy says. First is that, because of the new and “cutting-edge” nature of the technology, when they first try to use it, “they hit a block.” Often the “solution” turns out to be more expensive than anticipated, and the initiative founders.

Second, when working with vendors, shippers often attempt to implement AI themselves, despite their inexperience with the technology. In the course of doing so, they’ll focus on a specific and immediate problem, rather than the larger benefit of the investment. “They need an understanding of the entirety of what AI can do,” Nagaswamy says. “They need to get more education on that.”

Because AI is still in the experimental phase for so many users, they tend to smart small, he says. In future, he believes, they’ll spend more time on building out “the larger puzzle.”

AI agents are the hottest topic surrounding the technology at the moment. They are a smaller subset of generative AI, designed to address a particular action or set of workflows. At that relatively small scale, Nagaswamy says, there’s less of a chance that the AI will “hallucinate,” as the many manifestations of GenAI have been shown to do in their early days. As an example, agentic AI can monitor the flow of trailers to make sure that drayage is happening on time, and that units aren’t being delayed at ports and elsewhere, to the point where the shipper is being slapped with expensive detention and demurrage charges. D&D, he says, “can be reduced significantly.”

Still in its formative stage, AI has given rise to charges that it’s mostly hype at this point. Nagaswamy disagrees. “This is as real as it can get,” he says. “We have seen some very tangible outcomes.”