The future of automation – will machines ever completely replace people in the warehouse?
However, as we now enter the information age, we are beginning to see a new kind of invention – machine automation. Today, machines can land aircraft, diagnose cancer and trade stocks effectively, and this is just the beginning of their potential. This new age of automation could have a significant impact on the economy of the world with a 2013 study finding that almost half of all jobs in the US could potentially be automated in the next two decades.
Machinery in the warehouse is nothing new, but up until now, machines have always served to support a human component. However, with the advancement of automation, could we possibly see people free warehouse in the next twenty years?
What does automation mean?
We are all familiar with the idea that machinery can complete jobs faster than people. This is particularly prevalent in manufacturing, in the years between 2010 and 2016 manufacturing has seen 10 to 20 percent increases in output thanks to continually improving mechanical processes. However, these machines are simple. In car manufacture for example; machines like the Unimate perform a single function like welding and they do it over and over again. Unlike people, these machines don’t stop and often do jobs that could be considered dangerous.
Machines are currently excellent at performing simple, narrowly defined, predictable tasks, but complex jobs are still left up to people. However, if complex jobs are examined closely enough they can be split up into many narrowly defined predictable tasks, one after another. Machines are on the verge of being able to do these kinds of complex jobs very well.
What about in the warehouse?
The warehouse is no stranger to automation and a lot of the background processes are now done automatically. However, people still offer inherent benefits. Take picking for example; most modern warehouses will give their pickers scanners which will tell them the item they need to pick and require them to scan it. This is an automated process which works better with a person involved. A person can assess the quality of an item, knows where to find it on the shelf and can walk there.
Machines can’t complete tasks with this level of complexity yet, but this is gradually changing. Amazon Kiva robots, for example, pick up whole pallet racks and bring them to a station where a person then picks the item. This alone is said to save around $22 million per warehouse in fulfilment expenses.
This may sound like a simple process, but in reality, it’s quite complex. The Kiva robots need to know exactly where an item is on a shelf and calculate a route to the picker who needs it, all whilst communicating with other robots to avoid accidents.
What are the drawbacks?
Kiva robots work brilliantly for Amazon where the volumes warrant this kind of innovation, but for most companies, humans remain the cheaper option. Amazon paid an initial cost of $775 million to purchase around 30,000 machines. This works out at an average cost of around $25833 per machine and multiple machines are needed for each warehouse.
So will machines ever replace people?
Despite the advancement, the machines of today are still not able to complete complex tasks as well as a human worker. Amazon are creating more work for people with the Kiva system because physical picking still involves people. However, this is just one example of possible future innovations.
Machine processes are only likely to start replacing people on a mass scale once they can be implemented with minimal infrastructure development and this may be some way off. Current analysts predict that even by 2025 people will still be involved in warehousing. However, in the longer term, it is very feasible that machines will begin to reduce the number of people needed. Picking represents on average 70% of all labour costs in a warehouse, so huge costs savings could be made if machines do become more viable.
Click here to contact us, or subscribe to our mailing list by entering your email address at the bottom of this page.