Hitachi’s Robot Bosses

A Hitachi warehouse is being co-managed by artificial intelligence to increase productivity.

09.23.15 5:00 AM ET

Hitachi might be in the early pioneering days of eliminating one of the major causes of employee turnover: the manager.

The company instituted a program at one of its Japanese warehouses where an artificial intelligence system was integrated to help rearrange processes and find ways to induce more efficiency and productivity.

It’s a big deal when you consider the ever-complicating world of employee satisfaction, and the ever-diminishing separation between work and home. Bosses can create a lot of problems—after all, they’re only human. Letting an AI run endless calculations to find the best outcome certainly takes the hesitation out, and replaces worker hesitation with facts and numbers.

Warehouse workers in this particular Japanese location didn’t just come to work one day to find the shift manager made of titanium, speaking like a GPS system in an uncomfortable calm British accent. It worked a little more subtlety.

“The AI is closely linked with the business system,” a Hitachi spokesperson said. “The operation screen/dashboard used by the floor manager is the same as the previous business system with the addition of an ON/OFF button to apply or not apply AI.”

Basically this system, unlike a charming or insidious android, is just part of the faceless computer system. The manager (still human) gets to decide when to include or not include said AI in his decision-making process.

What’s more, the AI is only working to increase workflow by rearranging processes, not deconstruct them to look for wasted time. “The presumption is that each task is conducted independently,” said a spokesperson. “In this warehouse demonstration, the details of each task were not changed, just the sequence in which they are to be performed.”

For many workers it’s possible they weren’t even aware that there had a been a system change. “In the case of the warehouse where field demonstration was performed, there is no direct interaction between the workers and the AI system. The AI system introduced issues job orders in the same format as previously dispensed,” Hitachi explains, “and the job reports are submitted in the same way; the AI system ‘reads’ the performance data from the job report. The workers, therefore, may or may not even be aware of AI involvement.”

Plus, Hitachi says there was no sudden push for more efficiency by individual workers. “The sequence in which tasks were to be performed was re-arranged to relieve congestion,” he says. “The workers were not requested to work more or any faster, just perform the tasks in a different order so we do not envisage a situation of increased stress.”

It will take a while for the company to gather useful data on the experiment, but this project has the potential for a lot of impact as we learn more about the complex ways a human boss can negatively impact employees.

Managers likely hate being the bad guy when the top of the pyramid demands more productivity. And countless studies have shown that micro-managing done by supervisors can have negative impacts on more than just productivity, so it’s possible that integrating AI into the problem-solving arsenal will allow both of the human parties to communicate better—if only out of shared distrust for a robot overlord.

In the meantime though, factory and warehouse work will definitely benefit from efficiency increases that don’t demand more work from already productive workers. And managers probably are safe until such a time as the AI figures out a solution to the thermostat problem.