fbpx

Recommender Systems: from Customization to Personalization (and Beyond)

By November 20, 2021How Alberts Does It

What if you never got stuck on thinking about what to eat? What if someone else presented you day after day with the exact meal you did not even know you were craving?

Alberts is on a mission to make a healthy life the easiest option. How do we do that? By making healthy eating effortless. But, that sounds easier said than done, right?  That is we created the first automated blending vending machine for smoothies, soups and shakes with a personal touch. What does that mean exactly?

Customization is now mainstream

Today, consumers can already customize their smoothie. When you open the Alberts App, you can fill a cup with the ingredients of your choice. When you have crafted your delicious recipe, you scan the QR code to send the recipe to the robot. Our blending station will take care of everything else. 

This is basically what you can do in most restaurants as well today. You can pick and choose ingredients for your pizza. You can select side dishes to pimp up your meal. However, this does not allow the Chef in the kitchen to prepare something that is really personalized to you

alberts app

Our vision is to integrate an automated robot with an intelligent platform to go beyond just customization and create a real personal user experience. To make this happen, we joined the PERSFO.eu consortium with specialized partners who are experts in machine learning, data analytics, food service, etc.

How can we personalize on a large scale?

Personalized nutrition does sound like a utopian view on how the food industry and consumption could develop, especially when you envision a world where you can personalize to individual needs and preferences on a worldwide scale. 

So, how will we go about that? 

Let’s start with a situation most of us are familiar with. You’re finishing up at work and you get an email. It’s from Netflix. “We’ve added a new movie you might like.”

Now, how does Netflix know “what you might like”? Based on your preferences and view history, Netflix knows what you are most likely to appreciate next. Which is great, right? If you are not sure what you feel like watching, you will not lose too much time sifting through the list of movies, or worse; waste your evening by selecting a bad movie. 

This is what Alberts wants to do with food. We want to give you personalized recommendations for meals and snacks. Do you want to develop a balanced eating routine, but are not sure what would be good for you to eat? Our food recommender will make a personalized suggestion for you.

How does a recommender system work?

First our system will get to know you better. This way, the Alberts App can suggest the perfect smoothie recipes for you; just how you like it, and when you need it. 

These recommendations would include information about:

  • taste preferences
  • allergies
  • metabolic system
  • workout habits
  • location

Let us look at an example. Imagine a typical day when you just finished your workout at the gym. Your wearable or smartphone notifies you with a suggestion for your perfect smoothie to get re-energized. 

The platform will take into account: 

  • your preference for pineapple, 
  • your allergy for strawberries, 
  • your energy dip from the afternoon workout

The result: a smoothie recipe personalized to your taste and dietary needs.

Can we go beyond personalization?

The short answer: yes, we can.

As the science behind food personalization becomes more developed, there is a huge potential to take the recommender platform one step further.

Every person has a different DNA. From that DNA, doctors can tell if you are lactose intolerant for example. In a few years, your DNA information will help you make healthy food choices based on what you need as an individual.

The future of food is at our fingertips. Alberts is proud to be a pioneer in the food robotics market space and to collaborate with the other partners of the PERSFO project to make a healthy life the easiest for everyone.

Leave a Reply