At Cambridgeshire Climate Emergency (http://camemergency.org), we’re building a network of “Climate Emergency Control Centres” including a “situation map” (http://camemergency.org/map) that can be used to track progress and drive urgent action on zero carbon - ideally through regular daily, weekly and monthly goal-driven conference calls.
What we’d love to see in our situation map dashboard is granular, regularly updated C02 emissions per individual person that could then be aggregated through the normal Google Map zoom in/out methods (similar to what we do for concept mapping at climate.space). The challenge is, however, that things like individual energy meter data are not easily accessible (though we think we can get companies to maybe sign up to a voluntary “Make my energy public” system as some already do that).
So what we’re thinking is some kind of ML algorithm that takes a variety of readily measurable features, eg. socio-economic status, distance from conurbations, temperature, average traffic flow, time of year, etc, and calculates a very broad-brushed C02 emissions figure - that serves as a reliable proxy for actual, detailed C02 personal calculators (which we would use as well to calibrate the thing). That way people can see on a regular basis the “situation map” and change their behaviour - ideally collectively by creating locally grown energy schemes - that radically alter the situation map.