Machine learning for geographically differentiated climate change mitigation

Speaker: Felix Creutzig, TU - Berlin

Artificial intelligence is transforming many scientific disciplines but is only slowly adopted in research on climate solutions. A systematic understanding of the state of artificial intelligence and machine learning in research on climate mitigation is missing. It is also unclear, which area of climate mitigation research will benefit most from novel algorithmic infrastructures. Here, we perform a systematic bibliographic review of applied machine learning studies that are of relevance for climate change mitigation, focusing on spatial data and specifically on the fields of remote sensing, transport, and buildings. We find a relevant body of twenty years of literature that is exponentially growing. Our review identifies crucial avenues for upscaling solution-oriented research at a high spatial resolution, and for delivering globally consistent comparative policy solutions that respect geographical differences. We suggest a meta-algorithmic architecture and framework for using machine learning to optimize urban planning for accelerating, improving and transforming urban infrastructure provisioning.