Skip to content

Spontanets

This initiative arises as part of the inminent adoption of artificial intelligence methods for design and manufacturing industries. As commented in my previous blog post, companies are yet to capitalize on AI, and those engineers who embark on the journey of adopting deep learning methods for their processes will outpace the rest of the industry.

What is the vision behind Spontanets?

Spontanets is the place to both keep up informed of the latest trends in AI for Industrial design and manufacturing, and the platform where engineers can upskill their AI knowledge.

Currently, most of the engineers that participate in the shift towards AI adoption are PhDs with a deep knowledge of maths. However... - The population of PhDs is small compared to the total numbers of engineers, thus limiting the pace at which companies are adopting AI. - Most PhDs in the AI field for Physics are more worried about keeping up with the latest model architectures, rather than focusing on the real implementation in the industrial world. - There is no link between the Industrial development and the world of AI. Companies try to capture some of these trends by promoting internships with AI topics, but these projects tends to be left out without a real solution deployment.

I've personally experienced how hard it can be to get into this world. I am lucky to be part of a Research project within Airbus that aims at replacing expensive simulations with AI-based surrogate models, but not all people are exposed to these opportunities. Spontanets is oriented to exposing engineers to real problems and real implementations, so that they become ready lead the industrial transformation.

What are key ideas to have?

While speaking with a startup founder who's building an AI solution for Industrial Design, some words he mentioned sticked in my head: "The future is a prompt". If this is the case, the number of current processes/software (and the market share associated to them), will dramatically change. Most companies will die if they don't transition to AI, or will have to pay millions to those companies who win.

Potential upskilling roadmap

  • uv essentials: how to manage your python environments?
  • github: version control
  • introduction to statistical learning
  • how to look at your data?
  • surrogate models
  • advanced deep learning methods
  • introduction to transformers / generative ai