Ambi Robotics, a provider of AI-powered robotic sorting solutions for material handling operations, announced the launch of PRIME-1, the first Robotic Foundation Model to be deployed in commercial warehouse operations. PRIME-1, which stands for Production-Ready Industrial Manipulation Expert, provides a unified transformer backbone that can be fine-tuned for various robotic tasks, including 3D perception, package picking, and quality control. The model enhances performance, accelerates product development, and improves the reliability and maintainability of Ambi Robotics’ AI-powered robotic systems.
“With PRIME-1, we’re addressing the most pressing challenge in warehouse robotics: the need for adaptable, scalable solutions that evolve with operational demands and amplify return-on-investment,” said Jeff Mahler, Co-Founder and Chief Technology Officer at Ambi Robotics. “PRIME-1 allows our customers to leverage collective learning from our entire production fleet, empowering them to stay ahead in the rapidly evolving logistics landscape with increasing demand. Our customers now have the ability to respond faster to market dynamics and future-proof their operations in an industry where speed and precision are paramount.”
Training with Real-World Warehouse Data
PRIME-1 was pre-trained using self-supervised deep learning on a dataset of over 20 million high-quality images from individual pick, place, and pack events. These events were collected over 150,000 operating hours from Ambi Robotics’ fleet of AI-powered robotic sorting systems in U.S. warehouses. This dataset represents approximately 1% of the data collected to date, leveraging real-world warehouse operations data from scaled deployments to ensure precision and efficiency.
“The scale of training on real-world data collected from the AmbiSort A-Series systems has enabled us to reach high levels of reliability with PRIME-1,” said Vishal Satish, Foundation Model Lead at Ambi Robotics. “The use of PRIME-1 will allow us to rapidly develop and deploy new robotic solutions for a variety of tasks, while also improving the performance of our existing robotic systems.”
Advances in 3D Reasoning and Scalability
PRIME-1, trained on over 1 trillion tokens, has demonstrated the ability to generalize 3D reasoning for a range of tasks, including depth estimation and robotic picking. Production testing has shown clear scaling laws: as the volume of data used for pre-training increases, both pre-training quality and performance on downstream tasks improve. These findings highlight the potential of pre-training on large volumes of unlabeled data to achieve superior performance compared to models trained exclusively on labeled data.
“Emerging AI research shows that generative pretrained models can outperform previous architectures,” said Ken Goldberg, co-Founder and Chief Scientist at Ambi Robotics. “The engineering team at Ambi Robotics used 4 years of proprietary warehouse data to train a state-of-the-art generative model for 3D warehouse operations; their experiments with real production systems confirm that PRIME-1 significantly outperforms their previous systems.”
About Ambi Robotics
Ambi Robotics is an artificial intelligence (AI) and robotics company developing advanced solutions to scale ecommerce operations while empowering humans to handle more complex tasks. The company’s AI operating system, AmbiOS, leverages proprietary simulation-to-reality (Sim2Real) technology and the latest AI foundation models to provide reliable, high-speed robotic systems. Founded in 2018, Ambi Robotics is headquartered in Berkeley, Calif., and is home to a team of top roboticists, AI researchers, and industry professionals building cost-effective supply chain technologies.
For more information, please click here.
Source/Photo Credit: Ambi Robotics
(Editor’s Note: All trademarks mentioned in this article, including company names, product names, and logos, are the property of their respective owners. Use of these trademarks is for informational purposes only and does not imply any endorsement.)
