Nick Haase, co-founder of MaintainX, explains why artificial intelligence-driven automation initiatives can fall apart without a strong maintenance and data foundation.

The excitement over AI has some executives assuming that it offers an instant solution to efficient warehouse maintenance and operations. But in order for AI to achieve that objective, facilities need a strong data foundation. That’s especially true in warehousing, where critical intelligence doesn’t necessarily reside on the internet for instant access through applications such as ChatGPT.

When approaching the issue of data quality and availability, warehouses must start “at Day One,” Haase says. That’s because the data they have in hand “may not have enough contextual information for AI to do something interesting with it."

When rolling out AI, another big mistake that distribution facilities make is trying to overlay the technology on broken systems. Technology doesn’t provide the solution to that problem, Haase says. To ensure that they’re implementing the right procedures, warehouses need first to solicit the input of frontline workers, who must be closely observed in their day-to-day tasks in order to see where improvements need to be made. Operational executives, continuous-improvement experts and systems engineers should also be part of the teams that lays the groundwork for successful deployment of AI.

The ”first principle” of automation and digitization, Haase says, is to figure out which processes are working well and which aren’t, with a particular focus on the bottlenecks that are suboptimizing operations. That effort “is a great stepping stone to building foundational data for digital transformation.”

Will human workers be completely replaced by AI and automated systems in the warehouse? Distribution facilities “are at least 30 years away from being fully autonomous,” Haase says.