[LUM#22] I came, I saw, I conquered
The year is 2024 AD. Every sector has been taken over by artificial intelligence... Every sector? No! The stubborn industry continues to resist this revolution. A resistance that mathematician Bijan Mohammadi could well overcome with his magic formula.
He is neither a druid nor a warrior, but a researcher at the Alexander Grothendieck Institute in Montpellier. By swapping his cauldron for a computer and his potions for an algorithm, Bijan Mohammadi has developed an AI that is unique in the world and convinced manufacturers such as Renault of its relevance. "It uses physical signals such as noise or vibrations to detect anomalies and perform preventive maintenance, " he explains.
Hearing the tearing of a tiny rubber connector to move a defective car out of the way or detecting wear on a milling cutter head by monitoring changes in vibration—these are the types of applications this AI is capable of. Some might say that's no big deal, but... "It allows up to 300% more parts to be produced with the same milling head, representing a saving of several hundred thousand euros per year per production line," points out the mathematician , who is "accustomed to working in the industrial sector and its logic."
AI and industry: discord
Savings in money, time, and quality. That was all it took to convince this sector, because although we might imagine AI reigning over industry like Caesar over Gaul, that is not the case. There are many reasons for this reluctance, starting with the lack of robustness of traditional AI, whose fixed architectures are not designed to adapt to the variability of industrial sound environments. "We see this with voice assistants on phones, which malfunction if the user has a strong accent or if there is background noise," explains Bijan Mohammadi.
Added to this is the issue of data. For reasons of sovereignty, manufacturers refuse to turn to GAFAM to share the data needed to train AI. There remains the question of scaling up to industrial levels and the lack of proof of return on investment. In short, "Many difficulties, few guarantees—an equation that doesn't appeal to business," sums up the mathematician, who has therefore opted for a different formula.
From Caesar's laurels to Bondzai
In the summer of 2016 AD, while Gaul was turning in on itself, Bijan Mohammadi was working on his 12 labors and developing the algorithm that would open up the realm of the gods to him. The latter replaces the fixed architecture of traditional AI, which has a fixed and generally enormous number of neural layers, with a neuron generator that stacks layers in real time like a 3D printer. "It's a kind of miniature AI that creates its own architecture. The advantage is twofold: it needs very few examples and adapts completely to the environment in which it is placed. It is very reliable since it involves few specific parameters."
And with little data, there is no cloud, which means no confidentiality issues and no colossal energy costs. "It's a frugal AI that runs on conventional processors that require very little energy and no investment in IT infrastructure." Reliability , adaptability, profitability... These are the ingredients for success that have been marketed since 2021 by the start-up Bondzai, whose founder Bijan Mohammadi(x) continues to reap the rewards.
Read also:
Find UM podcasts now available on your favorite platform (Spotify, Deezer, Apple Podcasts, Amazon Music, etc.).