MHI Solutions Improving Supply Chain Performance
Collaborative robots (cobots) and autonomous vehicles are increasingly finding a home in warehouses and distribution centers (DCs). The success of those deployments, however, is not assured. Industry experts shared six best practices that can help achieve an effective implementation.
ISTOCK.COM/ONURDONGEL
Well entrenched in manufacturing processes, collaborative robots (cobots) and mobile autonomous vehicles (MAVs) have increasingly made inroads into warehousing and distribution center facilities over the past decade. As the cost of these technologies has come down, operations are leveraging them to enhance productivity and throughput amid ongoing labor shortfalls.
It’s in the best interest of operations planning to implement these technologies to make every effort to ensure their success, both in increased efficiency and acceptance by the existing workforce. MHI Solutions spoke with experts from four MHI member companies for insider details about how to do just that. Sharing insights are:
- Pete Allen, head of strategic partnerships at Agility Robotics;
- Josh Cloer, vice president of sales and marketing at Mujin;
- Dean Elkins, senior director of robotics and vision at Gray Solutions; and
- Kevin Heath, director of autonomous mobile robots at Dematic, a member of KION Group.
MHI Solutions compiled their recommendations into a list of six best practices that can increase the likelihood of an effective cobot and/or MAV deployment. First, however, here’s a quick review of these robotic technologies.
Defining Cobots and MAVs
Before diving into the best practices, it’s important to identify the distinctions between the two categories of technologies, as well as to introduce a third solution: humanoid robots. In the broadest respect, all three are robots. That is, programmable machines capable of automatically performing a complex series of actions. However, the similarities end there.
Cobots are multi‑axis robotic arms capable of movement in three to six different directions. Specifically designed to work alongside humans at a fixed location within shared workspaces, they incorporate safety sensors that trigger the cobot to slow or stop if an associate comes too close.
In warehouses, larger cobots have been most frequently applied to routine, repetitive tasks, such as palletizing and depalletizing. Smaller, more compact versions are also beginning to be used for eaches picking in order fulfillment. These smaller cobots’ programming leverages advanced vision systems and real‑time tactile feedback—coupled with artificial intelligence (AI) and machine learning (ML)—to autonomously adapt to item and placement variability.
According to Grand View Research, the U.S. market for cobots was $1.58 billion in 2023; by 2030 the firm projects a compound annual growth rate (CAGR) of 32%, to $11.04 billion. That forecast cites increasing adoption by small‑ and medium‑sized firms and anticipates significant deployment of pick‑and‑place cobots for order fulfillment.
This broad MAV category encompasses automated, autonomous stackers, forklifts and pallet trucks, as well as automatic guided vehicles (AGVs), which typically handle heavier, palletized loads, autonomous vehicles that tunnel under and move racks and shelving between storage and picking locations, and autonomous mobile robots (AMRs). Load capacities range from less than 220 pounds (100 kilograms) to in excess of 4,185 pounds (1,900 kilograms). The vast majority of MAVs are used as an alternative to fixed conveyors for flexible transport of loads and materials.
Different navigation technologies either guide the vehicles along a predetermined path or enable dynamic, autonomous travel. Depending on the type of safety sensors they sport and the process to which they are applied, MAVs may be segregated from their human colleagues by physical barriers, or they may travel and work alongside them, supporting various order fulfillment processes. In general, the heavier the load and the faster the travel speed, the more likely autonomous vehicles are to be separated from personnel. Because they often interact with humans, AMRs are sometimes referred to as collaborative.
Like cobots, the overall MAV market is also expected to grow considerably over the next few years. Research and Markets reports a combined market size of $5 billion in 2022, which it forecasts to reach approximately $20 billion by 2028. The firm projects CAGRs of 22% for AGVs and 37% for AMRs, resulting in a combined installation base in excess of 2.7 million by 2028.
The newest robotic solution making inroads into warehousing and distribution are humanoid robots, with a bipedal form factor that closely resembles humans. Their design allows them to be implemented in place of humans to perform repetitive, hard‑to‑fill jobs without requiring significant modifications to the facility or existing material handling equipment.
Last summer, GXO Logistics signed the first formal commercial deployment agreement for a fleet of humanoid robots. Supplied by Agility Robotics, the Digit bipedal mobile manipulation robot now works at a SPANX facility, integrating with existing systems and performing tasks such as transferring totes from AMRs to conveyors.
With the significant growth projections for cobots and MAVs—and the impending commercial expansion of additional humanoid robotic use cases—chances are good that more operations will be looking to benefit from these technologies. The following six best practices will help companies achieve those benefits as quickly as possible.
Best Practice #1: Engage an experienced partner.
As with any automation implementation, outside guidance from an experienced system integrator can help ensure a smooth deployment, particularly for the first—or first few—cobot or MAV integrations.
“Quite often we see customers whose innovation team decided to tackle a project like this on their own, but it ultimately winds up in a corner, collecting dust,” said Heath.
Most companies’ internal resources lack the expertise required to integrate robotics with existing automation and systems, he continued. These include the warehouse control system (WCS), programmable logic controllers (PLCs), warehouse management system (WMS) and any other complementary software and technologies.
“Working with an experienced system integrator can ease the pain of getting the entire system to work smoothly and successfully,” he said. “Getting that integration right the first time is super critical to success.”
Key to getting the integration right the first time is selecting a partner committed to fully understanding the operation and the workflow into which the robotics will be added, Cloer noted.
“Often, it’s the process that results in a deployment not being as successful as hoped. Therefore, it’s critical to have a good process before automating with cobots or autonomous vehicles,” he said. “Find an integrator who’s not just trying to sell you a piece of hardware, but instead will work to truly understand your process and can identify and help address the challenges associated with automating that process.”
For example, continued Cloer, a successful cobot deployment encompasses far more than simply moving a box from point A to point B. It also includes identifying the box’s size, shape and weight; ensuring the labels are facing a specific direction; understanding the different packaging types and grippers that work with each; parsing the assorted business rules restricting box placement; and more.
“An internal resource may understand the processes and limitations but may not have the expertise to build the system to think intelligently about how it’s going to solve those challenges,” he said. “Working with an experienced integrator can help to identify all of those ‘gotchas’ from the outset so they can be built into the cobot before it’s deployed.”
Best Practice #2: Match the robotic technology to both application and environment.
Although both cobots and MAVs are robotic solutions, each is best suited for different applications. Further, within each technology there are a variety of specifications, sizes, capacities and capabilities. That allows them to be tailored to the operation, the process and the environment.
For cobots, noted Elkins, a successful deployment hinges on ensuring that the payload capacity and associated operating considerations are adequate for the application—whether that’s palletizing or depalletizing, eaches picking, reading barcodes, inspecting picks or applying labels.
“Cobots are limited to a maximum speed of 1,000 millimeters (3.28 feet) per second in collaborative operation mode for the safety of the personnel around them,” he said. “Whereas an industrial robot might be two or three times faster. For the deployment to be successful, make sure the required eaches picks or case placements per minute align with the operational speed of the robot.”
Payload capacity and speed are likewise critical to the success of an autonomous vehicle deployment, continued Elkins. MAVs following associates up and down aisles to assist with order picking while navigating traffic have vastly different speed, load and safety parameters than ones segregated from other operations behind fixed protective guarding as they perform routine tasks.
“Environment is also an important consideration,” he added. “Some autonomous vehicles can be used in cold storage, but most are designed for dry facilities with fairly flat floor surfaces.”
Finally, clearance space for unimpeded travel is a necessity. “If they’re moving 1,500‑kilogram (3,300‑pound) pallet loads in two directions, autonomous vehicles need 15‑ to 18‑foot‑wide aisles in order to safely pass each other,” explained Elkins. “Smaller AMRs, or those only moving single file in one direction, don’t need nearly as much. But regardless of the size of the device, it’s important to confirm there’s enough available floor space.”
Best Practice #3: Conduct a risk assessment to ensure associate safety.
Many operations assume that cobots or MAVs can be used anywhere in a facility without segregation from personnel, but that’s not always the case, said Elkins. Instead, the determination of whether a robotic solution can safely operate among its human colleagues hinges on the findings of a risk assessment that also accounts for industry standard safety requirements.
“For example, as cobots have increased in payload capacity, they’re typically able to lift up to 35 kilograms (75 pounds),” Elkins said. “Depending on the tooling and the load, they may or may not be able to operate in the open. The findings of the risk assessment will help identify measures needed to keep personnel safe, such as equipping them with the right safety scanners to slow them down if people encroach on their work areas.”
Humanoid robots are often mistakenly assumed to be collaborative, added Allen, who said many manufacturers have created videos that do little to discourage that perception. However, as an emerging technology, there are currently no industry standard safety standards for bipedal robots.
“They’re dynamically powered, so if a humanoid robot loses power, it collapses. That creates a potentially dangerous situation for a human working next to it,” he explained. “The technology is certainly heading in that direction, with safety sensors that perceive people and objects, as well as safety stops. And safety requirements are being developed as well. But for now, they need to be physically separated from personnel.”