Ambi Robotics has launched AmbiStack, an AI-powered robotic system for warehouse automation. It stacks items onto pallets or into containers using Sim2Real reinforcement learning and AI vision. The system aims to improve efficiency, reduce costs, and enhance worker safety.
AmbiStack is designed for manufacturing, distribution, and ecommerce. It stacks mixed SKUs densely, lowering transportation costs. It also reduces repetitive tasks, allowing workers to shift to robot supervision roles.
CEO Jim Liefer said logistics companies face pressure to cut costs and speed up deliveries. He stated that AmbiStack automates stacking and palletizing with high accuracy and efficiency.
Unlike traditional automation, AmbiStack is flexible and adapts to different warehouse setups. It builds inbound pallets from trucks, assembles outbound pallets, and stacks in walled containers. Its design, based on gantry technology, allows it to fit into existing facilities.
AmbiStack’s AI vision system learns from 200,000 operational hours. It scans, tracks, and picks items while performing quality control. Its AI planning system reduces data collection needs, allowing fast deployment. Using reinforcement learning, it adapts in real time to optimize stacking.
Features include multi-container stacking, adaptive stacking based on item size and shape, and workflow flexibility. It supports both single-SKU and mixed-SKU cases. Multi-pick-and-place capabilities improve speed, and the system integrates with warehouse execution systems.
CTO Jeff Mahler said AmbiStack builds on Ambi Robotics’ AI technology. He expects it to improve over time, as seen with the company’s AmbiSort A-Series, which has improved 55% since launch.
Demand for warehouse automation is rising due to labor shortages and ecommerce growth. AmbiStack will begin customer deployments in mid-2025. Full availability is expected in early 2026 as production scales.