Other Voices: To get the most out of robots, optimize your people

Editor’s note: The following column by Justin Ritter, director of project engineering, and Ron May, principal engineer for Lucas Systems, is part of Modern’s Other Voices column, a series featuring ideas, opinions and insights from end-users, analysts, systems integrators and OEMs. Click here to learn about submitting a column for consideration.

————-

A 2019 report from ABI Research predicted that warehouses and distribution centers would install more than 4 million commercial robots by 2025 (this includes picking robots, autonomous AGVs, AS/RS and other types of automation). Much of the expected growth is for autonomous mobile robots (AMRs) that rely on human workers to pick and place products.

Like any emerging technology, growing adoption of AMRs will depend on buyers being able to cost-justify the robotic solutions for their operations. Paradoxically, a key to improving the business case for AMRs is to optimize and more fully coordinate the work performed by people alongside their robotic helpers.

Better optimization of human activities will reduce the number of robots needed per worker. Put another way, by optimizing the work of people working alongside robots, DCs will need fewer robots to achieve the same level of throughput.

Three categories of AMR
Since acquiring Kiva’s iconic orange robots in 2013, Amazon has installed tens of thousands of the goods-to-person AMRs in its fulfillment centers. Meanwhile, scores of new robotics companies have entered the market with a variety of AMRs that fit into three general categories: goods-to-person, robot-to-goods, and conveyance robots.

● Amazon robotics and other goods-to-person AMRs deliver shelves, boxes, or totes of products to workers at stations who pick the products.
● Robot-to-goods AMRs travel to pick locations where workers pick and place items on the robots.
● Conveyance robots function as take-away or transport systems to eliminate worker travel similar to conventional conveyor systems and automated guided vehicles (AGVs).

Conveyance robots represent a new twist on the challenge of optimizing and coordinating workers alongside traditional conveyor systems controlled by warehouse control systems (WCS). Robot-to-goods systems create an all-new optimization challenge.

Optimizing robot-to-goods by focusing on people
Robot-to-goods AMRs have shown the most promise in ecommerce DCs. They offer greater flexibility and a lower cost than traditional automation systems (including ASRS and similar goods-to-person robotics), especially for DCs in areas with very tight labor markets.

The robot control systems that coordinate and manage robot-to-goods systems do not directly manage the activities of the people working alongside the robots. In effect, the workers are ancillary to the robots.

In a typical robot-to-goods system, workers walk to the nearest AMR they see. When they arrive at the AMR they read pick instructions on the screen of a tablet mounted to the robot. After making the pick, scanning the item and placing it on the robot, the worker may then follow that robot or walk to another bot in the pick zone. This process creates dwell time at robots as workers read pick instructions on the screen of the robot. It also produces significant empty travel as workers walk to and from AMRs.

DCs can improve the productivity of workers and robots using work execution software and mobile applications to direct the work of pickers in coordination with the bots.

● The software can direct workers to make picks as they walk in the direction of their next AMR, reducing empty travel and increasing the pick density for workers (picks per unit of travel).
● Directing workers using mobile devices reduces time they spend reading the robot tablet. Workers can still confirm their pick by looking at the robot screen and/or by scanning the item at the robot.
● Having associates make picks independent of the bot reduces the time robots spend waiting in aisles and traveling within an aisle or zone. This increases the effective pick rate for the robots.

Work execution software can also manage the number of workers and bots in a work area as the volume of picks changes throughout a shift. If more picks become available, the system can direct additional workers and/or robots into the area and re-optimize the pick sequences and travel paths for the workers and robots. This is an especially valuable tool in today’s operating environment, where DCs need to consider crowding and spacing within DC aisles.

Published case studies indicate that typical robot-to-goods systems require three or more robots per picker. At those rates, AMRs may be cost-prohibitive for many picking applications. By optimizing the picking process for workers, work execution software increases pick rates and reduces the number of robots per worker. Reducing this ratio improves the cost-justification for robot-to-goods AMRs.

Optimizing conveyance robots
Conveyance AMRs offer a number of advantages over conveyors: lower costs, greater flexibility, and easier installation. Robots also don’t take up space in picking areas, and they eliminate the physical barriers that conveyors often create in a DC. Conveyance AMRs are a good alternative for DCs that could never justify the cost of conveyors, and for facilities that need to reconfigure or expand their storage and picking infrastructure.

As illustrated in these diagrams [see image], removing conveyors would allow this hypothetical facility to increase pick locations by about 33 percent. Without the conveyors, it is also possible to let workers move between areas that would otherwise be blocked off from each other.

The work optimization challenge with conveyance robots is very similar to that in DCs using conveyors controlled by a WCS. Work execution software is used to control the prioritization and distribution of work to pickers in different zones, and the pickers in the pick areas do their work as directed by mobile applications integrated with the work execution system. When pick assignments are completed, workers park completed carts or offload totes in staging locations for pick up by conveyance AMRs.

The work execution software coordinates the staging and takeaway of totes and carts with the robotic control system that directs the AMRs. This is analogous to the integration and coordination with warehouse control systems. Robots shuttle totes and/or carts between pick zones, and from picking to sorting or packing locations.

Bottom line
AMRs provide tremendous new opportunities to improve efficiency, but DC operators still need to think creatively about the best way to use the technology in their operations. DCs that optimize the work of people alongside robots can further reduce labor hours per unit shipped and improve robot utilization.