The allure of robotics in the materials handling arena needs no introduction. At modern retail distribution centers (DCs), where operations often run 24 hours a day, 7 days a week, 365 days a year, robots are increasingly being added to meet ever-increasing throughput demand. According to the Association for Advancing Automation (A3), North American companies bought 12,305 robots in the second quarter of 2022—a 25% increase over the second quarter of 2021 and a 6% increase over the first quarter of 2022. While the automotive industry has historically been the biggest adopter of robotics, recent years have seen a surge in implementations focused on automating warehouse logistics, according to A3.
In reality, the use of robots in retail DCs is not new. Many warehouses, for example, have used palletizing robots effectively for years to move heavy cases and position them in the exact same way over and over—which is an incredibly difficult task for people. And that’s just one example. For facilities that already depend on advanced automation—for example integrated, shuttle-based automated storage and retrieval systems (AS/RS); autonomous fork trucks; and high-speed conveyors within integrated goods-to-picker systems—the application of robots feels like a natural progression. Robots are just the next logical step in using automation to further address and solve operational challenges.
With these opportunities, however, come questions. What do those who manage DC operations need to keep in mind when implementing robots, and how should they approach the innovations in robotics that are now viable options for many warehouse operations? To start, DC operators should keep in mind the following six important considerations.
1. What is the difference between an integrated robotic solution and robotic point solution?
There are two main types of robotic solutions. Integrated robotic solutions are closely connected to existing automated systems, such as AS/RS, via software. Robotic point solutions work independently of those same systems. Integrated robotic solutions and robotic point solutions are suited for very different applications.
In most retail DCs that have a high level of throughput made possible by already existing automated systems, robotic point solutions have less applicability. That’s because these point solutions are harder to integrate into existing processes and workflows. However, robotic point solutions can be valuable for specific tasks, particularly in facilities that still rely on more manual operations. As a result, the common vision of robots/standalone machines that do everything—the very kinds of robots that dominate the headlines and often generate the greatest excitement outside of DCs—are typically of less utility in materials handling.
Some futurists may argue with this contention or note that fully robotic warehouses in which teams of robots move all materials are possible, but the reality today is different. Robotic point solutions can be used in discreet applications, such as stacking boxes that are always the same size and weight at a set frequency, with great effectiveness. For this reason, robotic point solutions are popular in environments like manufacturing where there is significant consistency—for example welding the same identical parts together repeatedly thousands of times. For now, however, robots are not able to address the innumerable ad hoc tasks that arise in distribution centers in the standalone fashion required for them to operate entire facilities on their own. In most retail DCs, the flow of products—the products being moved, picked, packaged, and shipped—is highly variable. Due to this, most retail robotics applications are and will remain integrated solutions. For example, AS/RS—which themselves can be considered robotic—are increasingly being paired with robotic pickers with exceptional success, both in increasing throughput and maintaining order accuracy.
Such systems are tightly integrated not only with one another, but to the DC’s core network that connects everything from when orders are received to when they are picked, packaged, delivered, and automatically restocked. In this way, the line between automation and robotics is blurring. Notably, these tightly integrated systems are possible because of recent developments in two key areas.
2. How have recent innovations made robotics more applicable to DCs?
Retail DCs historically faced issues applying robots because an order might include items of dramatically different shapes, sizes, or materials. Up until recently, most robots simply could not address such highly variable, inconsistent situations—let alone with the speed and efficiency of humans. Two advancements changed this dynamic and now play a key role in addressing the increased throughput demands that often prompt warehouses to deploy robots: improvements in vision recognition systems—the cameras and sensors used to help the robot “see”—and an increase in the array of end effectors, the “hands” that enable robots to manipulate or pick items.
With advanced vision recognition, robotic pickers now can see exactly which item they are tasked with picking—even shaking the tote or carton to make them fully visible or to determine which end effector to use.
Recent innovations in end effectors are no less noteworthy. There are now many variations of end effectors, from mechanical grippers to those that use compressed air for suction. As a result, a single picking robot can be used for a wider range of applications—even holding spare end effectors that it can swap in and out as needed to process different kinds of products. The applications of this kind of technology are virtually unlimited. For example, one robot can use an end effector with suction to move soft-sided packages, such as the bags often used for food items, and then quickly switch to a mechanical gripper for heavier, hard-sided items. Furthermore, now that robotic pickers have so many end effectors available to them, they can also be used to process larger and heavier items simply by increasing the scale of the robot itself.
3. What will be your throughput needs five years from now?
Many DCs struggle to determine how much capacity and throughput to address with robots, and whether they should look at peak periods for volume or at their average throughout. The more important question, however, is, “What will your throughput look like in five years?”
Yes, robots can be used as a short-term, or even a stopgap, solution to address very specific needs in the retail warehouse. Some organizations are even beginning to offer robots-as-a-service (RaaS), a subscription-based model that lets DCs lease robots during periods of increased volume or demand rather than buying them. It’s a viable option for some organizations that ensures that robots are only paid for when they are productively working.
However, a full robotic deployment, one in which robots are integrated into the core warehouse system and structure, is a much more advanced undertaking. To ensure that it is approached correctly, retailers should audit their operation to determine what throughput looks like as well as which products comprise most of it and which might challenge robotic systems.
Integrated systems often take two or three years to deploy from assessment to design, construction, testing, and deployment, so it is also imperative to look at least five years ahead to ensure that the desired return on investment is achieved. This does not mean that robotic systems should be configured to address periods of high throughput at all times, but rather that they are capable of being sped up to handle additional capacity on a seasonal basis or as needed.
4. What role should DC employees play in a robotics implementation?
Robotic systems have the potential to dramatically enrich the lives of DC employees, but those who run retail warehouses must actively include these employees in the selection process. Employees who currently work in roles shaped by manual, highly repetitive tasks have the most to gain from a robotics implementation. These individuals can transition into the highly skilled positions needed to manage, maintain, and repair robots, and their experience in roles such as picking can be exceptionally valuable when designing and refining robotic systems. For example, skilled pickers often have ideas on how and where existing picking processes can be streamlined and which kinds of effectors should be considered. Warehouse employees in general are also very aware of the many processes that would benefit from robotic automation. This information is of exceptional value when designing the layout of any distribution center.
Directors of retail distribution centers should involve these individuals in the planning process. Most importantly, it is imperative that they view robotic systems as an investment that complements warehouse staff and as a strategy that is proven to decrease employee churn. It is equally important to map out a career path for those employees whose roles will be automated or addressed by robots. Investing in and upskilling these employees is something that every organization investing in robots should examine.
5. How will you handle maintenance and repairs?
Robotic warehouse solutions are complex, and a breakdown can have a dramatic impact on operations. Once the decision to automate a DC with robotic systems is made, it is crucial to establish a strong maintenance and repair plan. This includes ensuring that adequate service level agreements are established and that the retailer has access to spare parts and experts.
One highly effective strategy for retailers is to build and hire their DC robotics team a full six months before the facility becomes operational. This allows the team to work side by side with the vendor in the testing and adjustment of all systems, simultaneously gaining familiarity with often expansive facilities.
Existing warehouse employees who are intent on learning about robotics and want to pursue a career in the field are ideal candidates for these assignments. Providing them with the opportunity to work directly with the vendor enables them to extract insights that have accumulated from many years of robotics work. Such opportunities should not be missed.
6. What tasks are robots still not well-suited for?
Enthusiasm for robotics is warranted. However, it is important for retailers to realistically appraise the capabilities robots bring to their facilities as well as which tasks robots are not yet ready to do at the velocity and scale required.
Two common tasks associated with the holiday shopping season offer a great example. While robots are proven in fulfillment operations, the reverse logistics involved during the returns season is problematic. While robots can scan returned items for volume and weight to determine if they are complete, today’s vision recognition systems are not ready for prime time use in the returns market.
For example, a robot can determine if both pairs of shoes from an order were returned and that they are the right kind, but it cannot determine if those same shoes were worn and can be sold as new. In the same way, a robot can determine if a returned television is the right model, but it cannot determine if it was dropped or damaged. People must still provide that critical oversight.
Vendors are working to quickly develop innovations to solve needs just like this. In the meantime, however, it’s important to leverage robots where they have the most potential for a return on investment.
Be thoughtful and realistic
It is also important to realize that robots are not suitable for every operation. Ultimately, what is important is not how many robots you used to address a problem but whether or not the problem was actually solved. Innovation for innovation’s sake just isn’t feasible for many brands, and not all organizations have the resources required to deploy large-scale robotic systems.
All retailers are, however, in a position to streamline their fulfillment operations so that they process orders more efficiently and accurately, retain valuable employees, increase the return on investment of their DCs, and address future periods when throughput needs are higher. Before they implement any automation, companies should make sure that it is based on best practices, done with reputable and experienced partners, and designed to address their own unique needs.
By keeping the above considerations in mind, those who oversee the operation of retail DCs can make sure that if they do move forward with a robotics strategy, they do so in a thoughtful and realistic way.
Those in our profession will likely look back on the present era as one shaped by unprecedented change and significant steps forward. Only time will tell if this is the beginning of the golden age of robotics or another chapter in the longstanding evolution of automation in retail DCs. What is certain is that those who embrace robotics thoughtfully, realistically, and methodically will see success in their fulfillment operations.