The AlphAI educational robot

Learning artificial intelligence with the AlphAI robot

In parallel to his activities in the Institute’s ICN department, Thomas Deneux has created the company Learning Robots, which develops and sells the AlphAI educational robot as its first product, to teach artificial intelligence (AI). This entrepreneurial project is supported and accompanied by the Université Paris-Saclay.
Thomas Deneux began his studies in mathematics at the ENS Paris, which he followed with a thesis in computer science on computer vision. Attracted by biology, he then decided to devote himself to neuroscience during a post-doctorate in neurobiology in Israel. « During my thesis, I analyzed brain imaging data. This made me want to learn more about the brain and practice experimental neurobiology myself » explains Thomas Deneux. « I then measured the activity of mouse neurons with imaging techniques. »

Mathematics at the service of neurobiology

With this experience behind him, he wants to put his mathematical knowledge to work in biology. « My greatest contributions are in the methodology field, so I wanted continue developing mathematical methods to serve neurobiology » says Thomas Deneux.
With this in mind, he joined the NeuroPSI Institute as a research engineer. His multidisciplinary approach to science led him to focus on the design of artificial intelligence algorithms. Drawing inspiration from neurobiology and animal behavior, he decided to develop a robot capable of assimilating tasks for which it was not initially programmed.

Tailor-made support to design

Like every year since 2013, the Université Paris-Saclay “Poc in labs” call for projects supports and accompanies innovative projects from the University’s laboratories with good potential for valorization. By granting funding, Poc in labs aims to develop the proof of concept of the project. It also offers project leaders support in studying the economic and legal potential of their project, and when necessary, an awareness of design orchestrated by the Design Spot. Finally, it offers a strategy for valorization and a team for the maturation and development of the project, and assists in the creation of a company thanks to training provided by IncubAlliance.
The design of AlphAI with a view to its commercialization benefited from such support and required the intervention of the Paris-Saclay Design Spot. It seemed obvious to us right away that a design study was necessary for this project, since it is a toy robot intended for a young audience,” says Vincent Créance, director of the Design Spot. First, we designed the product, the robot. Then, we worked on the logo of the start-up and the packaging of the product. Currently, we are already thinking about the design of future versions of the robot. “The entrepreneur and the designers decided to design a quasi-spherical robot and an arena made of pieces of wood that fit into each other, to multiply the game configurations.

Educating the younger generation about new technologies

Equipped with its various sensors (cameras, wheel movement detector), the AlphAI robot learns to understand its environment and to evolve in the arena. Depending on the learning mode chosen on the graphic interface, it becomes capable, through trial and error, of avoiding walls or following a track on the ground. The public then observes the robot’s behavior change over time and correlates it to the changes made on the graphical interface, which represents the neural network of the artificial intelligence. For the time being, the Learning Robots start-up, created in April 2020 with the support of the French Ministry of Education and Bpifrance, is demonstrating the robot in classrooms or at trade fairs. AlphAI is now being marketed in France and abroad, and the start-up hopes to expand rapidly.

Thanks to this playful and original approach to AI, Thomas Deneux aims to initiate future generations. “AI has become important in everyday life and will surely be taught in schools. It is necessary to develop new pedagogical tools now,” says the researcher-entrepreneur. Training these generations as early as possible in the main principles of AI will help them to better understand the technical and societal challenges.

Team leader
Thomas DENEUX – Research engineer