Using Machine Vision to Optimize Kitchen Operations
Intel RealSense Enables Real-Time Inventory Management for Food Service Automation.
Challenge
In the fast-paced food service industry, managing inventory can be a constant challenge. Overpreparing often leads to food waste while underpreparing can result in dissatisfied customers and lost revenue.
Solution
PreciTaste, a pioneer in AI-powered software automation, created Station Assistant to streamline kitchen operations and eliminate guesswork. This unique technology platform relies on Intel® RealSense™ stereo depth technology to detect, classify, and track food inventory levels.
Results
Station Assistant helps restaurants reduce costs, increase operational efficiency, improve labor productivity, and enhance customer satisfaction.
Introduction to Smart Kitchen Automation
Efficient inventory management is one of the biggest challenges aside from supply in the foodservice industry. Food waste is considered to constitute about 30-40% of the U.S. food supply, from which foodservice operations generate an estimated 12.8 million tons of food waste annually. Traditional inventory techniques are based on manual counting, which cannot be free from inaccuracies and inefficiency due to labor involvement. PreciTaste addresses those challenges with its AI-powered automation using Intel RealSense technology in optimizing food preparation and reducing overall waste.
Unveiling Food Service with the Power of AI
PreciTaste's AI-powered solutions have been observed to create measurable impacts along the value chain in food service. The reports indicate an 8% decrease in labor costs, 5% in food costs, and a rise in restaurant sales as high as 25%. These AI-driven tools will drive improved forecasting, predict demand, and optimize kitchen workflows, saving restaurant managers time to focus on improving customer experience.
Accurate Measurement Reduces Waste
Precision in inventory tracking helps minimize food waste. Intel RealSense cameras capture 3D dimensional data for measuring ingredients of variable sizes, shapes, and textures. Global shutters synchronize depth and RGB sensors for accurate food item detection, volume detection, and estimating the time of refills required. This stream of real-time data will guide kitchen workers, eliminating excess preparation and ensuring that ingredients will be ready when needed.
Rapid Development with the SDK 2.0
The engineers at PreciTaste applied the Intel RealSense SDK 2.0 to make the development and deployment of the system more efficient. It contains APIs for object tracking, 3D reconstruction, and spatial analysis that allow for quick sensor calibration and accurate modeling of food items. By fusing RGB and depth data, the system eliminates the need for complex separate solutions, saving time and reducing development costs.
Selecting Intel RealSense: A Vision for Smarter Kitchens
PreciTaste required a high-performance depth camera that could be seamlessly integrated with automation solutions. It had to combine RGB and depth sensing, precisely identifying the food items and estimating their volume. Since most installations in the kitchen would be mounted from above, it had to support long-range capabilities and clarity through glass displays. The Intel RealSense Depth Camera D455 did just that and was cost-effective to boot, reducing hardware costs by $500 to $1,000 per restaurant.
What's next?
Learn how RightHand Robotics is using Intel® RealSense™ computer vision technology to revolutionize automated warehouse order fulfillment.
Robotics overview
See how Intel® RealSense™ cameras can revolutionize the robotics industry with the latest in depth‑sensing technology.