New Kindred SORT 3.7 Drives Throughput for E-commerce Fulfillment with Expanded Putwall and Improved Material Handling

Kindred shares insights about AI robotics and reinforcement learning at MODEX 2020

MODEX 2020 - Kindred announced today it will debut the Kindred SORT 3.7 piece-picking robot at MODEX 2020 in Atlanta, March 9 to 12, 2020. The enhanced SORT 3.7 system features eight additional cubbies, for a total of 27, translating into a 42% increase in the number of items it can sort at e-commerce fulfillment centers and warehouses.

Co-Founder and Head of AI Research at Kindred, James Bergstra, will present at the MODEX Emerging Technologies Theater on Tuesday, March 10, 2020, at 11:30 a.m. He will share how Kindred utilizes reinforcement learning and closed-loop task control to command piece-picking robots to continuously make them smarter, faster and more accurate over time.

"Just like Kindred's AI and reinforcement algorithms are continually improving, the robotic engineering and features in the SORT are too," said Bergstra. "Together, these upgrades translate into increased throughput, speed and accuracy that drive e-tailers to meet their peak volume goals each day."

SORT 3.7 Increases Throughput with Induction Systems
The updated SORT 3.7 is highly-optimized to integrate with continuous induction systems, such as cross belt or bombay sorters, as it can now pick 400 units per hour consistently. When a Kindred SORT is paired with a primary sortation system, fulfillment centers can double the number of items one employee oversees and ships out. Collaborating with SORT, associates on average can ship up to 850 units per hour.

SORT piece-picking robots utilize AutoGrasp™, a robotics intelligence platform that identifies and selects items to pick and place. AutoGrasp combines vision, grasping and manipulation algorithms to handle clothing, poly bags and other small rigid items. Kindred uses cutting-edge AI research and human-in-the-loop data methodology to continuously improve robot capabilities so that picking becomes smarter, faster and more accurate over time.

For more information about Kindred and SORT 3.7, visit

About Kindred
Kindred is a robotics and artificial intelligence (AI) company that develops robots to solve real-world problems. Kindred's cutting-edge technology is the foundation of a number of proprietary platforms, including AutoGrasp™, developed to operate robots autonomously in dynamic environments. Its team of scientists, engineers, and business operators have set a new standard for reinforcement learning for robots. Kindred's flagship solution, SORT, is a piece-picking robot with human-like grasping capabilities that operates in retail environments to separate multi-SKU batches into customer orders. The company is co-located in San Francisco and Toronto. For more information, visit

Featured Product

OnLogic Karbon 430 Fanless Rugged Computer The K430

OnLogic Karbon 430 Fanless Rugged Computer The K430

OnLogic Karbon 430 Fanless Rugged Computer The K430 packs the power and advanced IoT capabilities of modern Intel processing into a low profile, highly customizable, rugged fanless system built for the challenges of the IoT Edge. Internal components are protected from dust, debris, chemicals, and moisture thanks to fanless and ventless cooling. The Karbon 430 features an operating temperature range of -40° to 70°C, 9~48 V power input, and zero moving parts, all of which help to ensure a long system lifespan.