Through UP2030, Istanbul aims to sharpen decision making to advance its planning agenda by initiating positive energy neighbourhoods with the use of advanced computational methods (i.e., digital energy twin and artificial intelligence) in support of decision-making. Artificial intelligence (machine learning methods) will be used to predict hourly building energy use and PV electricity generation, as well as urban mobility patterns in the selected neighbourhood. We will use deep learning methods, and we will specifically explore Recurrent Neural Networks, Graph Neural Networks and Multi-layer Perceptrons. The city will engage relevant stakeholders, from households in the selected neighbourhoods to decision-makers in IMM, through a serious game approach as an immersive and interactive Urban Building/Transport Energy Model (UBTEM). The emphasis will be:

• On the decarbonisation of buildings and transport systems and the effective integration of photovoltaic systems into building and urban public spaces.
• On transport, with emphasis on 100% renewables-fed e-bikes deployment, which will target inclusive use (note car ownership/use is male dominated: it is expected that e-bikes will give improved access to women, young, people with certain disabilities and elderly) .
• On measuring social benefits to reduce energy poverty and overheating-related health risks. 

Furthermore, the digital technologies embedded in the UBTEM will provide easy-to-explore, open-access data to citizens towards well informed decision-making for all stakeholders.