Robots already play a significant role in material handling. The modern-day world of e-commerce demands it. Amazon.com demonstrated its commitment to the technology back in 2012, when it paid $775m to acquire Kiva Systems. Renaming the unit Amazon Robotics, the company quickly began deploying robots within its ever-growing network of distribution centers. Amazon continues to invest in research and systems that will shave time off order fulfillment.
So does just about every other major e-commerce player, supported by countless vendors that are turning out robots capable of executing an increasing number of tasks, either working alongside humans or replacing them altogether. Still, the focus up to now has been largely on repetitive, transactional processes, in warehouses as well as on the production line.
More recently, we’ve seen a move toward robotics process automation (RPA), which isn’t just about swapping out humans for machines. It provides the “glue” that integrates multiple systems dedicated to order taking and fulfillment, according to Rodger Howell, principal with PwC’s Strategy& practice. RPA goes beyond physical systems to embrace the underlying software, with the help of artificial intelligence.
The potential market for deployment of RPA and accompanying systems is huge. PwC estimates that as much as 45 percent of current work activities can be automated, saving $2tr in annual wages. “In addition to the cost and efficiency advantages, RPA can take a business to the next level of productivity optimization,” the firm says.
The distribution center is especially ripe territory for use of RPA. The “lights-out” warehouse, where virtually every task is automated, has long been an option wherever labor costs are high, and speed of fulfillment is a priority. The latest technology ensures that those tasks are tightly coordinated, so as to create a smooth transition from placement of the original order to fulfillment and delivery to the customer.
Even where people are still present, RPA can perform a valuable service, says Howell. Four key elements need to be in synch: robots for picking and the movement of product through the facility, sensors that ensure product quality by detecting vibration and humidity, cognitive learning systems, and artificial intelligence, the last to turn processes into algorithms that guide the whole operation.
Putting the pieces together is still a challenge, says Howell. Companies must be able to tie their customer-relationship management systems back to order-management and enterprise resource planning platforms, looping in manual processes wherever they might occur. “It’s an attempt to get transactions automated in a much more structured way,” Howell says.
RPA isn’t something that you can just buy off a shelf. Multiple vendors need to be involved. “It’s a collection of integrators and different assets from the hardware and software sides coming together,” says Howell.
RPA is especially taking hold today in the area of procurement transaction processing, he says. Companies can use it to order discrete materials, without the need for investing in electronic data interchange for the sharing of messages. Other areas ripe for RPA include order management and data quality assurance. Without it, Howell says, companies might find it difficult to draw on product quality data from multiple systems, with some issues falling between the cracks.
The question of lost jobs continues to hang over the robotics debate. Howell believes human workers will transition to more creative jobs, as they have in response to past technological advances. “RPA is focused on automating, speeding up and simplifying transactional processes that many would find mundane,” he says.
Still, there’s a selling job to be done. RPA is a relatively new concept, and many companies continue to engage in debate over how far they should go in the direction of automation.
“RPA language is starting to get out there,” says Howell. “With almost any major business, you’re going to find pockets of this. The question is, are they doing this aggressively or not?”
The human equation has to be addressed. “Businesses are definitely going to take time to think about it, and they should,” says Howell. “It’s important stuff.”