Match Item Picking Robots to Handled Products for a Successful System Integration

When picking items, nothing is as dexterous, versatile, or adaptable as the human hand. That has made it exceptionally difficult to automate the picking part of order fulfillment. With nearly twice as many open jobs as there are unemployed Americans and exploding throughput demands, item picking robot designers and suppliers are responding with a host of new solutions. And that’s a good thing. Because there is significant variability among individual items. That, in turn, calls for greater variety in item picking robots.

Here, we explore the importance of matching item picking robots to handled products — plus a few other key factors — to achieve a successful system integration.

Product Variability Requires Sophisticated Item Picking Robot Control Algorithms

Products come in myriad shapes, sizes, weights, dimensions, and packaging types. And, as consumers have come to expect endless choices, warehouses must stock hundreds of thousands of individual items. Because of the sheer number of products, manually teaching an item picking robot to identify, recognize, and automatically handle each item is impossible.

Today’s item picking robots incorporate artificial intelligence (AI) and machine learning (ML) technologies. This eliminates the need to manually teach item picking robots the characteristics of individual products.

AI and ML — embedded within the robot’s control algorithms — enable the robot to determine how to pick an item it’s never seen before. This advanced capability allows item picking robots to accommodate packaging changes, the introduction of new products, and product diversity with a high degree of accuracy and reliability. It also enables the practical deployment of robotics to item pick applications.

Optimal Product Characteristics for Item Picking Robots

There are certain product characteristics that affect robotic item picking success. These factors include the packaging type, weight, and dimensions. For example, health and beauty products are often ideal candidates for item picking robots because of their relatively consistent size and weights. Conversely, traditional two-part shoeboxes are not optimal for robotic item picking, as the robot picks the lid but leaves the rest of the product behind.

Different Item Picking Robots Offer Different Capabilities

Different brands of item picking robots perform better with certain product ranges. Some systems are ideal for handling apparel, but not general merchandise — and vice versa. This means that an operation primarily fulfilling apparel orders will potentially choose a different robot vendor than a facility handling general merchandise.

In either case, the key to a successful implementation is properly managing the inflow of products. The robot should only receive items that is can handle.

Item Picking Robots as Part of an End-to-End System Solution

For item picking robots to function dependably, a reliable, finely tuned, fast, automated storage and retrieval system (AS/RS) equipped with goods-to-person (GtP) workstations must be in place prior to their implementation. Item picking robots pick at workstations located alongside those of their human colleagues, and their control software integrates with the operations’ overarching warehouse control software.

As directed by sophisticated software, the AS/RS directs totes of individual items to different picking robots configured to handle products with a unique set of characteristics. Products that don’t match the robots’ capabilities route to manual picking workstations.

With onboard AI and ML, the right mix of robotic solutions, and proper configuration, the number of orders that item picking robots can fill will steadily increase. That’s why it’s important to work with a vendor capable of integrating multiple types of robots from a broad range of suppliers into a complete, end-to-end solution.