For many fulfillment centers and warehouse operations, batch-picking is the most efficient process for fulfilling customer orders. Multiple multi-line and single-line orders can be picked at the same time, optimizing the picking process for SKU location, worker availability, and other factors. After orders are batch-picked, the picked SKUs have to be sorted out into individual orders and packed out for shipment.
A putwall, in combination with a put-to-light or similar system, is often used for this order sortation step. Each SKU in a mixed-SKU tote is scanned by a worker and placed into the specific cubby or bin location corresponding with a specific order.
During high volume periods or when staffing is low, manual order sortation using a putwall can become a bottleneck to the packing and shipping throughput. A robotic putwall can help reduce this risk by automating a critical step in the fulfillment process. This also applies to any returns processing operations currently using a manual putwall to sort SKUs to return to stock (read more about how automation can help your returns processing).
What is a robotic putwall?
A robotic putwall is an autonomous solution that uses AI-powered robots to pick items out of mixed-SKU totes, scan them, and place them into the corresponding cubby in the putwall. This helps ensure fulfillment throughput can keep up with surges in order volume, even when labor availability is low.
The ability of a robotic putwall to operate autonomously with high speed and high accuracy over long periods of time is crucial in determining how beneficial such a system can be to your operations. The key to driving that level of performance is modern deep-learning-based AI that is able to pick your SKUs without fail, regardless of how varied your SKUs are in shape, size, and packaging — and how often the SKUs turn over.
An AI-powered robotic putwall can help make your batch-picking and returns processing automation more efficient while reducing labor dependency.
Spotlight: Radial
Radial, a leading eCommerce fulfillment service provider, deployed 12 Covariant Robotic Putwalls at their Trade Port 2 facility. They’re using our AI-powered robots to sort millions of items for one of the largest retailers in the US.
Operating at 425 PPH and picking 100% of SKUs, the Covariant putwalls were a key factor in driving operational efficiency and throughput during peak season in 2022 in the face of labor availability challenges.
Key aspects of a robotic putwall
Piece-picking powered by AI
The sheer variety of SKUs in any fulfillment center can be staggering. Imagine a health & beauty fulfillment center: You might have tiny eyeliners and heavy jars of cream in the same tote as polybagged accessories and bunches of hair ties in a paper band. And these SKUs might change every season!
Modern deep-learning-based AI is necessary for a robot to be able to pick such a wide variety of SKUs. Universal AI, such as the Covariant Brain, can leverage vast learnings from millions of picks from connected robots in warehouses around the world. This enables robots to pick virtually any item on Day One. With fleet learning, all the connected robots, regardless of picking use-case, learn and improve from each other continuously.
Hardware designed for speed and accuracy
In order to maximize the benefits of AI, a robotic putwall should be designed to optimize speed and accuracy. Thoughtful solution design should take into account SKU variety, inbound and outbound tote management, and other upstream and downstream processes. One such example of innovative hardware design that maximizes speed and accuracy is the end effector or gripper used on the Covariant Robotic Putwall. Individually actuated suction cups of varying sizes allow the Covariant Brain-powered robot to use the same gripper to pick a wide variety of items with high precision and speed without dropping or double-picking items. Such hardware design is also used for other Covariant solutions for induction, goods-to-person picking, and kitting.
Seamless integration
In order to ensure order processing throughput is not adversely affected, it is imperative that any robotic putwall system implemented integrates into existing upstream and downstream processes. The robotic system should coordinate with the WMS and order management system seamlessly. Furthermore, the packer experience on the other side of the putwall needs to be consistent with any manual processes in place, with little to no training required. For example, if a pack-to-light system was used previously or is used in adjacent manual putwalls, the robotic system needs to integrate with that so the packer knows when to pack out orders without needing to learn a new process.
From an operational oversight perspective, order throughput data, such as the number of items sorted and a number of orders completed, must be integrated into any analytics and dashboards already being used. Additionally, there should be robot-specific operational overview that shows the breakdown of productivity and fleet status for each robot.
Robotic Putwall: Increasing warehouse efficiency
Whether you have a batch-picking operation or are processing a high volume of returns, sorting is a necessary but manual-intensive process. Labor availability and cost challenges can be mitigated with AI Robotics. Adopting AI-powered solutions, such as the Covariant Robotic Putwall, helps you increase operational efficiency and stay ahead of the competition.