Four tech directions to shape distribution center (DC) performance

Call it flexible automation or warehouse robotics, but by almost any measure, the highest growth categories in the materials handling world are autonomous mobile robots (AMRs) and smart pick-and-place robotics that make use of artificial intelligence (AI). But that doesn’t mean they’re the only solutions to consider.

The many robotic variants include AMRs that do assistive picking, AMRs that do case handling, and with robotic picking, some solutions autonomously pick eaches or polybags, while others can manipulate cases to automate palletizing or de-palletizing. For certain, robots are the biggest thing to hit warehouse automation in recent years, but the trick is to know which robotic solutions or combination of solutions match operational needs, while having the software to tie everything together.

The other reality to keep in mind is most warehouses have a mix of systems, automated and manual, and most rely on labor—warehouse associates—to get orders out the door, even when robotics allow for smaller teams to do more.

Warehouses also have constant inflows and outflows of inventory that need to be received, stored and shipped accurately and in a way that minimizes costs. These needs have given rise to solutions aimed at coordinating robots with people, and in tapping vision technology and AI to understand inventory flows.

Here are four trends to watch. All involve robotics or AI, and could shape the way more warehouses or plants with complex materials handling can amp up operational performance.

1. Orchestration platforms

Wearables scanner provider ProGlove sees its integration capabilities as an enabler of effective human and robotics collaboration.

A sign of just how prevalent robotics has become is that now integrating different robot solutions with each other and with traditional automation, like conveyors or automated storage and retrieval systems (AS/RS), has become a software solution space. Analyst company Gartner calls this segment multi-agent orchestration or “MAO” platforms.

Dwight Klappich, research vice president and fellow with Gartner, says the need for MAO platforms is driven by companies that already have robotics and are not just looking to scale up one solution, but get multiple ones working together, while being able to integrate them with existing automation. They want the type of tight integration that can be had with traditional systems integration projects, but less hard-coded and rigid.

“They want to be able to bring a different robotics solution in and get them to work together as a holistic system, but they want to maintain that flexibility and adaptability,” Klappich says. “They want integration and orchestration, but they don’t want to be locked in for the next 20 years.”

The MAO platform market is still emerging, Klappich explains, with differing vendor capabilities, though most have a mix of integration mechanisms and business logic to set up workflows that span systems.

“Basically, it’s a layer of software that sits above heterogenous fleets of robots and other forms of automation to orchestrate work and integrate to underlying platforms. If you look at some of the big third-party logistics providers, they’re aggressively adopting robotics, and they are some of the companies at the forefront of this space saying, ‘hey, we’ve got to figure out to connect all these things more easily.’" - Klappich

Orchestration platforms provide integration ease but also fulfillment logic to manage workflows, says Akash Gupta, co-founder and CEO of GreyOrange, which offers an orchestration platform called GreyMatter. The orchestration engine assigns multiple agents to drive work across various areas and material flows. The software also uses AI to see patterns and learn how to make workflows more efficient, while providing dashboards on performance.

The platform can be used to integrate distinct systems into a cohesive flow—such as linking a robotic picking system with sortation AMRs, Gupta says, but there’s also interest in blending different types of robotics within a single process, like order picking, to arrive at a more efficient solution.

“Let’s say a fulfillment center’s order mix needs items that come from racks transported by AMRs, but some other items are in totes,” Gupta says. “With our platform, you could basically look at creating a hybrid station that can combine both racks and totes together, rather than a more complex consolidation process.”

GreyMatter is more than middleware, Gupta adds, because it provides metrics and insights to operations teams. However, Gupta adds, speeding up integration and onboarding of new systems is also crucial. “You can bring in new technologies very easily,” he says. “And, a year later, if you want to bring in another solution, you can pretty much do it right away without changing any of the existing integration, which makes the automation journey much easier.”

As more companies deploy more robotic systems, expect the need for orchestration software to accelerate, agrees John Santagate, vice president of robotics for supply chain software company Körber, which has partnerships with robotics vendors and orchestration capabilities. Robotics hardware innovation continues, he believes, but the end game should be operational improvement.

“The orchestration applications enable interoperable fleets, allowing multiple robotic solutions to execute with awareness of one another, either within the same workflow or simply in the same operating space,” Santagate says. “With this, the appropriate work can be sent to the appropriate robotic solution area, refining operations as we know it.”

2. Labor plus bots

Gupta notes that the agent concept means the orchestration platform also can tie into people-centric systems like voice solutions or pick-to-cart processes. “Robots and people are basically agents that register into GreyMatter and get guidance on how to execute processes and when to execute,” he explains.

Various other solution vendors, some of them with deep roots in robotics, are also touting solutions that eliminate clunky handoffs between robotics and human-centered workflows. Wearables can play a key role here, say vendors.

Paul Travers, president and CEO of Vuzix, says its augmented reality (AR) glasses are well suited to tasks like picking cases to pallets, either on a manual pallet jack or to a pallet on an AMR, because they keep the worker hands-free and can present visual instructions from pallet building software on the glasses.

“Glasses are truly hands free, and they also can provide a visual representation of a process like a pallet build, which helps avoid problems like crushing of lighter products or an unstable load,” Travers says. “It provides a visual representation of exactly where everything needs to go, which supports faster, more accurate builds. That visual element also helps with onboarding.”

OX, a Vuzix partner, offers Vuzix glasses as a wearable for its software, a frontline operations platform that uses AI-enabled workflows to speed up processes like case picking with a pallet jack. According to Tanner Green, OX’s chief product officer, the OX solution is aimed at digitizing and optimizing manual processes, though tapping AMRs for transport is possible.

ProGlove, a provider of smart wearable scanners that fit on a glove, sees its software integration capabilities as supporting rapid creation of workflows that blend human associates and robotics. Ilhan Kolko, ProGlove’s chief product officer, says ProGlove has put intense development effort into its integration layer with AMRs, industrial robots, cobots and other forms of automation.

“We’ve been really growing our connectivity and integration portfolio over the past 18 months, involving 32 software releases, so we can bridge that [human to robot] gap. By enriching our integration portfolio, this ensures it’s easy to have our hardware on the worker, and our software running smoothly together with our partners’ software, to make integrated human and machine workflows happen for the customer.” -Kolko

ProGlove has integrated its wearable scanning with solutions from AMR partners including Locus Robotics, Geek+, iFollow and others, adds Kolko. For instance, ProGlove’s software simplified the integration between its wearable scanners with a goods-to-person AMR fleet from Geek+ at a BMW Brilliance automotive plant in China.

Metadata, like times between scans, can also be shared with robotics solutions to provide a more holistic view of productivity and traffic patterns, adds Kolko. “Our job is to make sure we have an open platform so anyone who has a mix of human resources and automation they want to bring together, they can easily do it,” he says.

At Onward Robotics, formerly known as IAM Robotics, its “Meet Me” solution focuses on coordinating mobile robots with human pickers as independent workflows, using wearable devices from major device vendors to notify workers where to go and what to do next, says Lance VandenBrook, CEO of Onward Robotics. He explains that while AMRs do provide efficiencies versus unoptimized manual picking, time can be wasted in coordination between humans and AMR fleets, which he says the Meet Me system addresses.

Onward Robotics offers AMRs with put-to-light shelving on top, similar in look to some other AMRs. The difference with Meet Me, VandenBrook explains, is that the solution’s algorithms maximize productivity by orchestrating human and AMRs workflows independently, while the system leverages wearables to guide associates on tasks and keep them aligned with the robots. The overall solution provides team-based and individual performance views.

VandenBrook adds that third-party logistics providers (3PLs) are especially interested in solutions that blend AMRs with human pickers in a way that supports pay-for-performance, which he says was a guiding principle for the Meet Me solution.

“You’ve got associates who maybe aren’t enabled to be as time efficient as they could be; and meanwhile in the market overall, operations are starved for labor resources, which is why we want to maximize every second of human resource available,” says VandenBrook.

3. Vision & inventory

Vimaan’s computer vision and AI technology captures shipment condition and quantity insights in addition to bar code data.

Another area of innovation is the use of computer vision technology and AI applied to inventory tracking and visibility to avoid problems like stock outs in forward pick areas.

If a facility relies only on manual bar code scanning to track inventory and feed data into systems like warehouse management systems (WMS), what typically happens is errors, unless the operation is willing to scan each and every pallet and case coming in while visually inspecting for issues like crushed cartons or cases, says S.K. “KG” Ganapathi, CEO and founder of Vimaan, which offers computer vision and AI solutions aimed at 100% inventory accuracy.

Vimaan’s solution set makes use of computer vision to automatically capture data on inventory at key points in a warehouse, including at receiving docks, in putaway, for cycle counting, or in packing and shipping areas. The computer vision and other sensors Vimaan uses can be deployed in a portal setup at dock doors or on hardware that allows the vision technology to be attached to lift trucks.

The gear doesn’t have to be everywhere, just at select points where key inventory events naturally take place, says Ganapathi, in a way that blends the data capture with natural materials flow.

This stream of data feeds into Vimaan’s AI-enabled software and continuous learning model that makes sense of the data to detect issues like pallets that are short on quantity, or damaged goods, explains Ganapathi, in a way bar coding can’t, to avoid problems like shortages.

“The costs of not having that true visibility and understanding what’s being received and other inventory events, is actually very high,” Ganapathi says.

Ganapathi adds that while it’s possible to achieve adequate inventory control through rigorous bar code scanning, human inspection and manual quality checks, such brute force carries an operational cost. “Much of the base ROI for what we offer comes from the fact that we will reduce the number of people required for these inventory quality control tasks,” he says.

About 20 sites are using Vimaan’s solutions, which provides a digital twin view into where inventory sits in a warehouse. Other vendors also are applying computer vision and AI to inventory visibility, including Kargo, which applies its solution at dock doors.

4. Robotic frontiers

Robotics innovations still matter, even with advances in other areas, though with robotics the software capabilities often set vendors apart. Overall, there continue to be more types of AMR solutions, such as sortation AMRs or autonomous case handling AMRs. Other frontiers exist, with vendors making progress in areas like in-aisle autonomous picking (see box), or humanoid robots from startups like Agility Robotics or Apptronik.

As Santagate observes, some use cases remain relatively untapped for robotic applications. “Currently, most AMRs are indoor applications, leaving immense opportunity for outdoor mobile robotic applications, such as yard management or shuttling between buildings,” he says.

However, Santagate adds, outdoor robotics faces environmental complexities that crop up outside the four walls. “In the building, you have a structured and semi-static environment, offering a mapped-out area where the robots can navigate with a sense of confidence and understanding. When you leave the building, the environment is more subject to change and therefore unpredictable for robots. This makes the ability to do localization far different.”

Santagate also expects to see more uptake for autonomous lift trucks, given the many workflows manual trucks are used for, and immense needs around pallet lifting and handling. “Every warehouse worldwide leverages these technologies, from fork trucks to stackers, but there’s still room for innovation, and I suspect this sector will be the next growth lever.”