by Michael Wolf

Whether it’s to carry groceries around the store or to deliver them to our front door, it won’t be too much longer before everyday shoppers see robots both in and around the grocery store.

But one potential interesting new use-case for in-store robotics we haven’t heard much about is for detection of produce freshness. That may change soon, as Simbe Robotics, the maker of the Tally 3.0 robot, has just been issued a patent for spectral imaging of produce and meats and detect how fresh they are.

The US patent, which is number 11,200,537 and titled “Method for tracking and characterizing perishable goods in a store,” uses computer vision to record images across a period of time and derive a set of characteristics specific to the type of food. For produce, it can assign a percentage of ripeness, determine whether it is under, over, or at peak ripeness, and determine if there is other biological matter such as a contaminant on the food. It can also determine whether a fruit or vegetable is rotten, damaged, or bruised. 

Figure from Simbe Robot’s New Patent

From the patent:

The computer system can access and implement hyper-spectral template histograms or template spectral profiles for “underripe by three days,” “underripe by two days,” “underripe by one day,” “ripe,” “overripe by one day,” “overripe by two days,” “spoiled or rotten”, and “moldy” for specific varietals of fruits and vegetables or for fruits and/or vegetables generally. Similarly, the computer system can access and implement hyper-spectral template histograms or template spectral profiles for “fresh,” “rancid,” “low-fat,” “moderate-fat,” “high-fat,” “low-water content,” “moderate-water content,” and “high-water content” for specific varietals of meats or for meats generally.

For those unfamiliar with the Tally 3.0 robot, the company first unveiled its latest in-store mobile grocery robot in October of 2020. The robot, which wanders grocery store aisles to monitor product levels and detects misplaced items, utilizes computer vision and AI algorithms to capture and provide data to store managers more quickly without needing to send as much information to Simbe’s cloud platform. 

This type of mobile inventory checking technology is valuable enough, so much so that grocers like Schnucks have already started deploying the robot across the entire chain of stores. Others, like Hy-vee, are in trials with the Tally 3.0 and likely will expand their fleets over time. 

As Simbe’s robots add the capability to help grocers fight food waste – one of the most significant cost drivers for the notoriously thin profit margins in the grocery business – chances are we’ll see more grocers adopt in-store inventory robots en masse.