Explore our R&D with the world’s most dynamic humanoid robotRead More
Discover the past innovations that informed our current productsRead More
Meet the team behind the innovationsRead More
Learn how we develop and deploy robots to tackle industry’s toughest challengesRead More
Start your journey at Boston DynamicsRead More
Stay up to date with what we’ve been working onRead More
Discover the principles that guide our work and policiesRead More
Warehouse Robotics
Blogs •
On the surface, automated depalletizing looks relatively straightforward. A computer vision system identifies a box and then a robotic arm with a gripper lifts the box off a pallet and onto a conveyor belt. However, solving carton location and path planning to successfully unpack a pallet requires highly advanced perception and intelligent vision processing algorithms.
Computer vision is nothing new, but many existing solutions fall short when faced with imperfect conditions. Inconsistent lighting, high variability, and high volumes of items—all common in warehouses and other logistics environments—will confuse the average vision system. Central to this problem is that most machine vision systems were developed for a manufacturing setting, used to identify the same objects in the same conditions over and over again. Think picking a component from a set location, moving it to a conveyor, and repeating indefinitely. The component, location, and conveyor are all the same every time.
By contrast, the warehousing industry thrives on variety. Every day thousands of pallets move through thousands of warehouses and DCs. Some pallets are single SKU, some are rainbow SKU, some are mixed SKU. Some have banded boxes or damaged boxes or items that aren’t boxes at all.
Building on our expertise in autonomy and manipulation, Boston Dynamics has introduced a more accurate and intelligent machine vision solution designed to enable fast, effective warehouse automation.
Only a purpose-built, machine learning computer vision solution can keep up with all the variation in the typical inbound warehouse environment with messy pallets, thousands of SKUs, and irregular cases. Boston Dynamics designed a computer vision system that functions in real-world warehouses and distribution centers.
This vision system uses machine learning algorithms to enable robotic depalletizing of both single and mixed-SKU pallets with minimal to no prior system training. With both high-resolution 3D and 2D sensing, the system is able to accurately locate a wide variety of boxes in challenging environments, including low light.
Not only can our vision system recognize and pick a variety of cartons in a variety of conditions—tightly packed single SKU, highly graphical, banded or wrapped boxes, packages of bottles and cans—it can also identify peripheral items like empty pallets and slip sheets.
With sub-second image processing, the computer vision system is extremely fast, minimizing robot dwell times and maximizing pick rates. This helps to eliminate bottlenecks in your inbound operations, ensuring newly arrived cartons are processed quickly so shipments don’t stack up.
And, perhaps most importantly, the system is highly accurate. As a result, operators spend less time simply ensuring that the automated system works as intended. Even complex situations, like overlapping or out-of-place boxes, don’t cause errors. Fewer errors means fewer operator interventions—and ultimately fewer disruptions to the flow of goods through the warehouse.
A purpose-built computer vision solution not only improves the overall efficiency of the depalletization process, but also allows employees’ effort and focus to be redirected to higher-value activities. Manual depalletizing is a slow, repetitive, and physically taxing process. Even with an automated layer picking solution, operators still have to contend with high error rates, unpickable items like shrink-wrapped trays, and the need to manually dispose of slip sheets. Best case scenario, manual depalletizing is a boring and frustrating task; worst case scenario, it leads to repetitive stress and strain injuries.
Labor is one of the highest costs in the typical warehouse, while high turnover and labor shortages are among the most common operational challenges. Automating repetitive activities and freeing employees up for other work makes warehouse operations safer and more rewarding for employees. All the while, improving retention and productivity, and ultimately driving ROI. To learn more, contact Stretch sales today.
Recent Blogs
•3 min read
How Flexibility Simplifies Warehouse Automation
•4 min read
Break Through Supply Chain Blocks
Tackle Trailer Unloading Woes