Artificial intelligence-based prescriptive analytics enable a network approach to planning in volatile times, says Joe Bellini, executive vice president of product management at OneNetwork.

Pandemic-inspired uncertainty resulted in buffers of inventory and other excesses. “What we have now is the worst situation of all, where the plans have all that excess investment in them, working capital and so forth,” Bellini says. “And we still can't execute to them because of all the disruption.”

The solution, he says, is to design the execution piece so all the moving parts are aware of each other. “We can actually do a multi-party, multi-tier model of our network. We can do it in real time, all using the same data. And therefore, as things begin to shift and change, we can react to it in real time, rather than have to push things into planning and hope for the best, which hasn't worked out so far, as everybody sees.”

In addition to the challenges brought on by the pandemic, there’s opportunity as well, Bellini says. Companies were forced to operate with lower inventories of the mix and volume they wanted to deliver to market. “They said, that's bad in a way, because we're not able to service to the level we want to, but in another way, we've had to learn how to operate with less inventory in our networks,” says Bellini. “So as we return to normal, they say, what should we look like in the future? Wouldn't it be great if we could continue to operate in this sort of make-to-order world that we all got somewhat used to during the pandemic with limited supply? Is there a way to operate more like that permanently? Can we actually increase our customer service levels but also decrease the working capital required to hit those service levels along with the labor [required]?”