“This is quality without compromise” – Moley Robotics on the chefs of the future

As artificial intelligence and cloud computing technology develops, the use of automation and robotics in the food sector promises to become increasingly advanced. For UK-based Moley Robotics, that means revolutionising the kitchen.  

Founded in 2014 by CEO Dr. Mark Oleynik, the company partnered with German robotics company Schunk to create two robotic hands capable of recreating restaurant-quality recipes designed by top chefs. Meals on the robots’ database include soups, paellas, steaks, omelettes and pasta dishes. 

Earlier this year, Moley Robotics opened a showroom in London. Just Food caught up Dr. Oleynik to hear how he believes the technology could change commercial cooking. 

Dr. Mark Oleynik: ​​​​​​​Since 2014, we have been developing different types of robotic kitchens. It’s been a long journey because this is a new product on the market: we invented this product and developed it from the scratch. There were no other references on the market which could help us to learn, so we became the pioneers.  

The main value proposition of our robotic technology is that you can save time. You get a high-quality product and different type of dishes, originally branded from different well-known professional chefs. In commercial kitchens, this is a labour-saving technology. It can replace a number of human operations during the standard cooking process. Our machine can help to optimise workflows and help enterprises to increase and improve the quality of food because of the absence of human mistakes and errors. It is also accurate in timing, following a sequence of cooking operations to prepare a particular dish.  

Dr. Mark Oleynik: We patented algorithms which help the robot to react to changes in the cooking environment in real time, because if something happens during cooking, the robot needs to understand this and needs to react accordingly. We have already been granted various patents in this area and we are now one of the leading companies who own their robotic cooking patents across the world.

Dr. Mark Oleynik: The challenges are around making the machine functional and safe. When you make a machine which is designed to do something specific, it can fail in some particular areas. The point is to keep this failure below a required threshold, so we put a 99.7% success threshold on the execution of each operation. This means that if there are more than 20 operations, around 95% are getting a positive result in the cooking process. However, this is our minimum range; of course, the maximum is 100% and we try to be on the maximum side.  

We do a lot of tests and work a lot on training the machine so that it won’t cause significant harm to a human, even one doing stupid things, which we understand happens; humans sometimes do strange, illogical things. That’s why we try to train our robot and design the machine in a way that won’t ever cause significant harm and there are standards for that. We follow the standards and we pass the tests which we need to pass in order to say that our machine – this collaborative robot – that we have in our kitchen is safe. 

Dr. Mark Oleynik: For us, the success factor is measured by whether the robot can cook the same dish as a human, or at least substantially the same. It’s definitely not better because it’s a robot recreating a recipe.  

I don’t think it will be possible for robots to cook better than humans in the next many, many decades. Even with all of the technology we have on the market today, it’s impossible to measure the taste of a dish, so we cannot say “this dish is tastier than that one”. There is no sensor and no information from a scientific point of view. Unfortunately, only the original chef can taste the dish and say whether they are satisfied that it is the same. That is what we do when we record a new recipe and we teach the robot how to cook it. When we have finished processing and generating the data, the chef who created the original dish comes to our lab, to verify and see how the robot cooks and then they taste the final dish. When they are absolutely happy with the quality, consistency, taste and smell, we can publish the recipe inside the database. I don’t think any other algorithms could be available for decades. 

Timing is also important because restaurants have to preserve some meal components when there isn’t time to cook a complex dish whilst customers wait. For example, if you cook risotto, you need to spend an hour cooking it and nobody can do it quicker. Chefs might cook half of it and then preserve it to be finished when ordered; however, there is compromise on quality. With Moley’s robots, if you want your food at 2pm, the robot calculates when it needs to start cooking and can start at 12:45. This is quality without compromise.

Dr. Mark Oleynik: No, I don’t think kitchens will be fully robotic. It really depends on the particular meal – if it’s a standardised menu, with X number of dishes, then it is already possible to run this type of cooking process completely with robots.  

We can see this to some degree in industrial manufacturing and food manufacturing because we generally produce a lot of food using machines. When you buy something from the grocery store or the supermarket, a lot of the food is made by machines. From pasta to sausages: everything is made by machines. Generally, production is fully automated, as people try to make it more efficient and more standardised. 

That’s why I think the next question is “can we manage a restaurant like a factory?” Yes, we can do it, but the problem with this is that, if it is very standardised, you probably won’t be able to change menu because it will be connected to an exact process. That could be bad news: if you produce millions of packs of pasta, then you are producing it for a market which is usually huge and you can sell it around the world. If you produce a meal, it needs to be eaten ideally right after you produce it. That gives some limitations because you want a freshly cooked meal of maximum quality. If you preserve the meal – maybe you freeze it or try to keep it hot for a while – the quality is going to go down dramatically. You get reheated food and this is a different quality, of course.  

Dr. Mark Oleynik: The main motivational factor for me is that I love nice food from different chefs. I don’t want to eat the same burger, sandwich or salad, day after day for my whole life; I want to have a good variety of dishes. I want to taste different cultures but I also want to keep to my dietary requirements and keep my body safe and healthy at the same time. That is why I started to create Moley robots, because I understand that the value of this technology will grow as more recipes are added.

Dr. Mark Oleynik: The future is pretty simple. Any new technology passes through three big stages. In the first stage, everybody says it is impossible. In the second stage, everybody says that it is possible but it is very complicated and expensive. In the third stage, everybody knows about it and says it is simple. I think we are between the second and third stages. Things are still expensive compared to cheaper kitchens, although compared to luxury kitchens, we are actually on the same price range. Usually, the companies who sell the big kitchen brands invest a lot of money into marketing and branding; we are using the same money for our robotic technology, but we are not doing so much in branding. We try to make sure that, in the kitchen, people get the maximum they can get for their money. That gives us the chance to sell robotic kitchens in the same price range as other big brands who sell their kitchen without.  

However, it is not enough. At the beginning of the Moley journey, I said that one day our kitchens would cost between £15,000 ($19,668) and £25,000. We started from £250,000 to £300,000; now our simplest kitchen costs around £50,000 to £70,000. It will be a big step forward to make it £15,000 to £25,000 unfortunately because it requires mass manufacturing which requires big numbers of units but that’s our next goal over the next five to ten years. For now, I think we have a nice, sexy product, with good functionality, a good value proposition, well designed, and I think we are on the right track.  

In general, in the future, I hope that one day everybody can have the nicest ingredients and can grab a nice dish from Tim Anderson (or another big chef) and can eat good quality, freshly cooked meals.