Doughnut economics provides a framework for appraising how to improve social behaviour through new relations and goals in the way humanity uses (and suffers from the lack of) resources for services. Municipalities used to have several competences and responsibilities either in the use of resources and the provision of (social) services to citizens. Usually they hold a huge amount of data and experiences that can be used to provide better planning of their responsibilities to citizens. Cities abstract social behaviour for regeneration and distribution in two aspects: i) those related to the interaction of individuals inside big communities, basically cities, and the global beneficials obtained by social services and solidarity, and ii) the social relationship of these communities (cities and citizens) with nature, in the sense of intertwining with sustainability and reasonable use of (natural) resources. In [9], Dahl proposed and justified Doughnut Computing, an AI methodology based on inferential programming that, using the Doughnut Economics model as a compass [21]), can help us achieve, both locally and globally, the goal of enabling social and ecological well-being. In this paper, we discuss how to specialize this methodology, always guided by the human and sustainability goals of Doughnut economics, to analize cities' social behaviour as communities in order to perform planning of municipal services to improve social relations, sustainability and quality of life. This paper reports our first experiences in three different cities across the world, showing a very promising interplay between computation, sustainability, and social services.
