Welcome to the food robot roundup, where we look at some of the most interesting stories in food robotics from the last couple of weeks.
Last month, Beewise, a robotic beehive startup founded in 2018, announced an $80 million Series C funding round and unveiled a new version of its Beehome system. Beehome is an outdoor box that functions as a solar-powered automated apiary. It enables climate and humidity control, pest monitoring, swarm prevention, automated harvesting, and real-time problem alerts to combat colony collapse disorder.
Beewise Using Robots to Rescue Distressed Bee Colonies
Colony collapse disorder is an abnormal phenomenon where adult honeybees suddenly disappear from the hives, and the queen and young bees are left unattended in the hive. This leads to colony collapse because the queen and young bees cannot survive without the adult honeybee. Scientists have not identified a single cause of CCD and have thus concluded that it is a complex combination of multiple factors including pathogens (disease-causing organisms).
Beewise claims to reduce bee mortality by 80%, increase yields by at least 50%, and eliminate 90% of manual labor compared to traditional beehives. They offer their product to farmers through a robotics-as-a-service model costing $400 monthly plus $2,000 for delivery/setup fee that includes 24 colonies and ongoing maintenance. Farmers are willing to pay for the robotic beehive because it improves bee yields and the pollination of surrounding plants. Beewise has raised $120m in funding to date and has saved over 160 million bees in the past year. The startup currently manages more than 7 billion bees, equating to 25,000 acres of pollinated crops.
Beewise isn’t the only startup trying to save the bees. In October of 2021, BeeHero announced $19M in new funding to scale its markets and find more uses for its active honeybee hive data collection. The startup uses IoT tech to track the movements and health of bees in real-time and has worked with several of the top almond growers in California.
FlyTrex Lands in Texas
Flytrex drone delivery is making its first U.S. expansion out of North Carolina to Granbury, Texas (right outside DFW) as part of the $40 million Series C funding round it raised last November. The startup is promising restaurant takeout delivery in under 5 minutes to those who live within one nautical mile. It will achieve this through its partnership with Brinker International, a restaurant group with brands like Chili’s and Maggiano.
Flytrex targets suburbs as opposed to urban areas because it can be expensive to do delivery on the ground in areas where people and restaurants are more spread out. The challenge for Flytrex, though, is to get FAA approval to expand the range of delivery past one nautical mile.
The people behind Flytrex’s new city expansion and fleets of drone delivery are trained “flight operators” who work via third-party companies to operate the drones. In fact, remote operators seem to be a growing segment of the workforce as robotic delivery expands. Even some autonomous delivery robots on the ground aren’t fully autonomous and rely on people on the ground to pilot them.
Fruits of Labor
We discussed the challenges of automating fruit picking, especially when the fruit is small and delicate, in the last food robot roundup. And while U.S. startups seem to be struggling to make rapid progress, researchers and engineers in other countries have been tackling the same world.
Researchers in Japan have developed a dual-armed machine capable of peeling a banana. Although this might sound like a very simple task to a human, it’s impressive that the robot can do this without bruising the fruit inside. The researchers used a deep imitation process where they demonstrated the banana-peeling action hundreds of times to the robot to produce sufficient data for the robot to learn the actions and replicate it. It took more than 13 hours of training for the robot to learn and it is only successful 57% of the time. However, there’s potential for this to alleviate the labor shortage problems Japan is currently facing in the food processing industry.
At EPFL, a Swiss university specializing in natural sciences and engineering, engineers are training robots to understand the process of picking raspberries. Delicate fruits like raspberries can be expensive for both consumers and farmers because harvesting the delicate fruits is difficult. High labor costs and a shortage of workers also contribute to the high price to harvest and resulting loss of harvest.
“It’s an exciting dilemma for us as robotics engineers. The raspberry harvesting season is so short, and the fruit is so valuable, that wasting them simply isn’t an option,” says Josie Hughes, a professor at CREATE Lab.
The team has designed and built a silicone raspberry that can tell the robot how much pressure is being applied. This is measured by a fluidic sensor consisting of a soft silicone tube within the raspberry. The raspberry separates from a 3-D printed receptacle that models the stem and it is attached to the raspberry by two magnets. The silicone properties can be adjusted to simulate the fruit’s resistance and require the robot to exert the necessary picking force. The feedback the raspberry provides trains the robot to harvest the fruit without damaging it. In addition to applying the right force, the robot also needs to learn how to loosen its grip once the raspberry detaches from the receptacle
Now what does the future look (quite literally) like? The next step is for the CREATE team to develop a camera system that would enable the robots to “see” where the raspberries are located and identify whether they’re ready to be harvested. The team is also hoping to apply this process to other fruits that are difficult to harvest such as berries, tomatoes, apricots, or grapes.
You can watch the raspberry-picking robot in action below.
Sodexo Expands Reach of Tech-Forward Solutions at School
U.S. foodservice contractor Sodexo announced plans to increase the convenience of its offerings to college students through virtual brands, ramen vending machines, and check-out free grocery stores. These three digital initiatives aimed at college students will accomplish Sodexo’s goal of engaging a younger, more technology-savvy audience that values convenience, affordability, and variety.
Students will be able to order from virtual brands through Sodexo’s partnership with Virtual Dining Concepts, which offers brands like Mr. Beast Burger, Mariah’s Cookies, and Buddy V’s Cake slice. Virtual brands are already offered by Sodexo at Gonzaga University but will be expanded to more campuses, although it’s unclear which ones. Adding to the convenience is the feature that meals will be delivered via robots.
Sodexo also announced a partnership with AiFi, an artificial intelligence firm, to offer checkout-free grocery shopping at the University of Denver starting this month.
The third offering of ramen and udon vending machines will be accomplished through Sodexo’s partnership with Yo-Kai Express, an autonomous vending machine manufacturer that is already present on some campuses.
College students can be a difficult customer group to capture because many are on a dining plan. However, Sodexo believes its offerings can be a complement to meal plans, allowing students can get meals after dining halls close. In fact, Sodexo has been making a variety of plans to target college students including announcing they would bring 1,000 delivery bots to college campuses in partnership with Kiwibot and expanding Jamba by Blendid to more campuses.
Beyond college campus foodservice, Sodexo also provides catering, facilities management, employee benefits, and personal home services to various sectors, including hospitals, senior living communities, government agencies, military bases, and prisons. Some of these settings might also be conducive to automation so it will be interesting to see if Sodexo graduates its use of robotics beyond schools.