research · CO₂ · deep learning

Deep learning models
of CO₂ chains.

Predicting the carbon emissions of objects in supply chains. Virtual twins of physical objects.

uncircularity model

living with earth

The largest problem we face is the world we share.

🌍 Earth's changing environment is perhaps one of the largest problems we face. Although human relations and economics shaped our world, we now have new tools. Machine learning and AI became paths to finding solutions — and the results from models now affect us, and our world.

make models

One of the first CO₂ deep-learning systems.

🥦 We move toward building models which understand and reveal our physical relationship with Earth. The uncircularity team built one of the first CO₂ deep-learning systems to predict emissions of products and global supply chains. Models show us the present and help us plan the future.

Open the worksite

model charts

learn

Larger, deeper, more physical.

Future models will be larger and more detailed. They will encode not only CO₂ but also other physical characteristics — energy flows, molecular dynamics, the texture of a real economy.

See the slides XOR graph