Cities-Board: A Framework to Automate the Development of Smart Cities Dashboards

Smart cities' authorities use graphic dashboards to visualize and analyze important information on cities, citizens, institutions, and their interactions. This information supports various decision-making processes that affect citizens' quality of life. Cities across the world have similar, if not the same, functional and nonfunctional requirements to develop their dashboards. Software developers will face the same challenges and they are likely to provide similar solutions for each developed city dashboard. Moreover, the development of these dashboards implies a significant investment in terms of human and financial resources from cities. The automation of the development of smart cities dashboards is feasible as these visualization systems will have common requirements between cities. This article introduces cities-board, a framework to automate the development of smart cities dashboards based on model-driven engineering. Cities-board proposes a graphic domain-specific language (DSL) that allows the creation of dashboard models with concepts that are closer to city authorities. Cities-board transforms these dashboards models to functional code artifacts by using model-to-model (M2M) and model-to-text (M2T) transformations. We evaluate cities-board by measuring the generation time, and the quality of the generated code under different models configurations. Results show the strengths and weaknesses of cities-board compared against a generic code generation tool.