Most warehouses today are operating systems provided by multiple vendors. Akash Gupta, chief executive officer of GreyOrange, talks about the need for orchestration of all those applications.

“Stage two” of warehouse automation and robotics finds facilities shifting from figuring out how a particular piece of technology works in isolation, to how it creates “end-to-end value for the customer,” Gupta says. Automating just one part of the process “is just shifting the bottleneck from one place to another.”

Humans, robots and underlying technology must come together, beyond the notion of systems integration. What’s needed now, Gupta says, is “orchestration” — ensuring that all pieces of the puzzle are operating in harmony.

Artificial intelligence, of course, is key to the success of warehouse automation. But the large language models that are getting so much attention today are just part of the picture, Gupta says. They need to combine with more “traditional” AI, in the form of machine learning, so that the larger system can make “more contextual” decisions.

Within the next five to 10 years, virtually every aspect of the warehouse will be automated, across the core elements of mobility, manipulation, sensing and documentation. With the help of AI, facilities will be able to assess manage multiple capabilities, and react more quickly to ever-changing market needs, Gupta says.