Moley the Robotic Chef: The Future of Cooking or An Expensive Toy?
Moley Robotics has launched a robotic chef that can make dishes resembling the skills of a MasterChef. Is Moley starting a revolution in the kitchen or is it an expensive robot with limited usability?
Choose the dish you crave and let your robotic chef do the magic for you, including clearing up and washing the dishes afterwards. That’s what Moley is designed to do. The futuristic kitchen was designed by London based Moley Robotics, founded in 2015, and is promised to be launched for customers in 2018 [1]. The introduction of Moley falls within a wider growing trend of household robotics that has been seeing a compound annual growth rate of 31% in 2017. The International Federation of Robotics estimates that more than 6.1 million units of household robots were sold in 2017, valued at $1.6 billion, marking a one third increase of those sold in the previous year [2]. Can Moley harnessed this trend through an economical, innovative and sustainable product or is this just an expensive toy that will soon fade away as the company garbles in managing this sophisticated machine to make a practical product with an appealing value proposition?
Capitalizing on recent improvements in machine learning and robotics, Moley Robotics sought to meet an increase in demand in household’s robotics by building cutting edge intellectual kitchen. The two-armed robot learned cooking by observing MasterChef Tim Anderson, the champion of the UK’s MasterChef TV show in 2011, through wired gloves and specialized 3D cameras [3]. The special kitchen which sits behind a clear glass also includes a stove, a sink and a counter where the robotic hands operate to create the desired dish. Beyond, the convince of a machine prepared food, utilizing the applications of Artificial Intelligence (AI) and machine learning in food preparation has the potential to minimize food waste by optimizing recipes from available ingredients at the time [4]. However, at $15,000 and with limited track record about its reliability the futuristic kitchen remains to be a hard sell for most customers [5].
Betting on the fast development of machine learning capabilities to mimic human movements, the smart kitchen company hopes to take a leap in making cooking human-less soon through supervised learning. The company is currently working to build a library of recipes, taught by the actions of human chefs, to reach a critical mass of dishes to appeal to customers with the right appetite of food and risk. In the long run however, the bet is placed on the machine’s ability to learn, unsupervised, making improvised decisions that require judgment closer to the way a human would make. This is easier said than done as we are still in the early stages of teaching robots supervised multi-dimensional skills let alone unsupervised skills and deep learning.
The challenge of machine learning is very complex and requires cutting edge innovation and a large sum of capital to make meaningful progress. On October 2018, Rethink Robotics, one of the leading companies in developing cost effective robots that can be taught new skills, shut down its operation after running out of money and failing to find a buyer. Rethink Robotics has developed two robots, Baxter and Sawyer with the latter learning its move by simply having a human manually teaching it the required moves [6][7]. The striking similarity between Rethink robots and Moley, with some arguing the superiority of the former raises serious questions about Moley Robotics ability to make money. The company will need to secure enough capital to keep it solvent and ride the wave of personal robotics demand growth to monetize its innovation. Moreover, the Moley kitchen will need to quickly pass the point of being a very expensive toy that can only copies the movements of a human to a truly smart machine that can take the household task to a new level. Only by rapidly developing its ability to perform its task in real world set up with minimal human intervention can it make a compelling story for its prospective. But even then, a serious value proposition to consumers in terms of improved convince, at high reliability and relatively low cost, will need to be made to make the expensive robot worth the investment.
This leaves a key question, will Moley be the nucleus of the kitchen of the future or is it a matter of time before it closes its door and follows the steps of Rethink Robotics?
Word Count: 703
[1] Moley. (2018). Moley – The world’s first robotic kitchen. [online] Available at: http://www.moley.com/ [Accessed 12 Nov. 2018].
[2] “Executive Summary World Robotics 2018 Service Robots”. 2018. International Federation Of Robotics. https://ifr.org/downloads/press2018/Executive_Summary_WR_Service_Robots_2018.pdf.
[3] Going, Claire. 2018. “Introducing: Moley— The World’s First Robotic Kitchen | IST 110: Introduction To Information Sciences And Technology”. Penn State University. https://sites.psu.edu/ist110pursel/2016/10/23/introducing-moley-the-worlds-first-robotic-kitchen/.
[4] Willis, Angelica, Elbert Lin, and Brian Zhang. 2018. “Forage: Optimizing Food Use With Machine Learning Generated Recipes”. Stanford University. http://cs229.stanford.edu/proj2017/final-reports/5244233.pdf.
[5] Gibson, Megan. 2018. “Meet The Robot Chef That Can Prepare Your Dinner”. Time. http://time.com/3819525/robot-chef-moley-robotics/.
[6] Bray, Hiawatha. 2018. “Robot Pioneer Rethink Shuts Down”. The Boston Globe. https://www.bostonglobe.com/business/2018/10/03/robot-pioneer-rethink-shuts-down/NlzXXX6NimgyDZYao0TlfO/story.html.
[7] Guizzo, Erico. 2018. “Rethink Robotics’ Sawyer Goes On Sale, Rodney Brooks Says ‘There May Be More Robots'”. IEEE Spectrum: Technology, Engineering, And Science News. https://spectrum.ieee.org/automaton/robotics/industrial-robots/rethink-robotics-sawyer-robots.
Thanks Energy for sharing this. This strikes me as an extremely interesting application of ML with the classic question of whether the robot will be able to succeed without human intervention. In this case, I would wonder what it would take for a robot to handle variability such as running out of clean dishes or ingredients. It doesn’t appear as if the ML has reached the point at which the robot can begin learning how to handle these types of situations, without instructions directly being programmed in the specific situation.
The other question I have is which market this robot would target – and different markets raise significant challenges. If the robot is tailored to the high-end restaurant market, I wonder how it could achieve the ‘art’ of the best human chefs. If the robot is tailored to a wider population, I wonder whether it would ever be cheap enough in the medium-term to justify use in a cafe or a home.
I feel this has a lot of potential for more lower-end restaurants, where people really don’t care much about the quality of what they are eating, but more about getting it fast, cheap and with lots of fat or sugar in it. In the higher-end restaurants though, I am afraid customers would feel like paying for a much lesser experience than that they have with an human chef. Part of the fun of going to different restaurants is actually experiencing what different chefs have to offer and this could give out a “standardized” feel anywhere it is used.
This case is super interesting – thanks for sharing!
Reading this reminded me of things like the “Robot Restaurant” in Tokyo, and I could absolutely see people going to a restaurant to see a robot cook their food. I have a tough time seeing it becoming mainstream in the home with such a high price tag, but as more home automation arises and prices for this technology falls I could see a machine chef certainly becoming a mainstay in every house. I’d be very curious to see how meal feedback could be incorporated for the robotic chef to improve upon existing recipes and potentially create new ones through machine learning.
Personally I think it might be a bit ahead of its time, but would 100% go to the robot’s restaurant if it had one.