SAN FRANCISCO: Chef Robotics, the creator of ChefOS, the AI Brain for food manipulation, announced today that its ChefOS Enabled Robots in production have achieved a major milestone as of January 2024: more than 10M meals in production. This is more meals in production than any other intelligent food robotics company in production.

Powered by ChefOS, the company’s flexible robots are helping food companies increase production volume to meet high demand and maximize revenue—all amid an unprecedented labor shortage that has overwhelmed the food industry (according to the Bureau for Labor Statistics, there were 1,137,000 jobs unfilled in food preparation and service in 2023). The Chef Robotics automation solution is also helping keep the American food supply chain onshore (as many are thinking of off-shoring given labor shortages).

In real-world production deployments at customer sites, Chef robots automate complex food assembly tasks by deftly manipulating a vast variety of ingredients and successfully mimicking the dexterity of humans. In the past, achieving this level of automation encountered one primary barrier: a lack of readily accessible training data to achieve human-level dexterity.

Unlike cloud-based AIs like LLMs, one cannot download the internet for training data. The training data has to be created and given that each customer has different ways of cooking, prepping, and combining ingredients that change every so slightly day by day, the training data has to be continually collected onsite at customer sites via real-world deployments.

Chef’s training data has grown exponentially since its first deployment in 2022. Both the dexterity and training data allow ChefOS-enabled robots to have high precision inference and manipulate the ingredient while being consistent, not damaging it, not spilling it, and also being able to manipulate various portion sizes. Consider that it took Chef 314 days to get to 1M meals in production, 100 days to reach 2M, an additional 21 days to get to 3M, and only 18 additional days to get to 4M. This rapid acceleration of data collection and learning is speeding up the time-to-value for new deployments and the ROI for every Chef customer. Chef has now manipulated hundreds of ingredients in production.

Based on strong demand from new and existing customers, Chef expects to more than triple its fleet of AI-enabled robots in 2024 to serve US and Canada customers.

Fonte: Business Wire