ODI Leeds

#AirHack Challenges

After a successful and thought-provoking open meeting, where more than 100 people across three separate sites joined in the discussion, we have gone through all of the notes and suggestions made in the Defra Data open document. Whilst there is an absolute deluge of different and interesting topics that could be tackled, we have narrowed down the challenges to cover air quality. The rest will have to wait for another time ;)

Smarter Behaviour, Cleaner Air

'How do we encourage user behaviour to change in relation to air pollution?' There are many contributory factors in air quality, but perhaps the biggest is emissions from transport. A busy network inevitably causes hotspots, where excess traffic causes more emissions. Can routes be planned better to relieve those hotspots? Could regular road users be encouraged to use alternative transport? Is it time for extreme measures - should cars and other vehicles be held to account for their emissions?

Data-powered Citizens

'How do we engage local citizens and encourage them to take an active part in the environment around them?' With open data, everyone can be involved. Instead of relying on data scientists based in local councils or organisations, could local people help crunch numbers, gather data, make changes, tell data stories? What should we be measuring? What are we actually breathing everyday?

Better Air, Better You

Living near heavy traffic increases dementia risk - The Guardian

Air pollution damaging your mental health - New Statesman

Stories like those above broke within the first month of a new year, and London famously broke it's yearly emissions limit in the first WEEK. What are the effects to our health and can they be tracked? And what could be done about it? Who can or should be helping?

Modelling - The Naked Truth

A majority of air quality predictions are actually modelled, based on various sources of information such as weather forecasts, emissions factors, etc. But these can be influenced by mistakes in those sources. A glaring example - the emissions scandal that hit several car manufacturers. Anything modelled using the scam emissions factors would be far from what was really happening. So how do we strip these models bare? How do we get accurate and honest modelling? Would the use of real time sensors and reporting fix the issue of inaccuracies? If yes, then how do we get it all to work together?

These challenges are by no means prescriptive - if you have an idea, please do get in touch! :)