Byggkollen AI - An analysis tool for Construction projects
Purpose and goal
Isolve is planning to develop an AI that can help identify potential problems and bottle necks in both small and large construction projects. Early know ledge that a construction project has the potential to have problems gives the possibility to remedy these problems early on w hich can generate large savings. Construction projects are very complex and need to consider a significant number of variables that is difficult for one person to go through. An AI is therefore well suited in controlling and analyzing relevant variables as w ell as recommending potential solutions.
Expected results and effects
Isolve is planning to deliver an AI that can be helpful when planning, handling and performing a construction project. This AI will be delivered to customers and partners of Isolve through Isolves product Byggkollen and is expected to be delivered in a first version by the end of the project.
Planned approach and implementation
Isolves AI work will be done in work packages and will consist of: WP1: Planning WP2: Identification/preprocessing of data WP3: Development of model(s) WP4: Testing/evaluation of model(s) WP5: Inkcorporation of model(s) in product The project will use the tools: Pandas, Scikit-learn for analysis and TensorFlow for modeling of the neural networks.