Using AI for Planning Predictions - Development of a Data Enhancement Engine

The construction industry is prone to cost overruns, time overruns, and quality issues. These are all caused by several factors, among which improper planning resulting from unreliable assumptions and unpredictable risks plays a significant role. This could be overcome if the planning-related issues are well predicted, and the risks are accounted for appropriately. Prediction of the various aspects of planning could be enabled by identifying the trend in past data and making it useful for current and future planning. Also, the data being semantic is significant to perform decisive predictions. This could be done by leveraging growing technology like Machine Learning, and Natural Language Processing which are all a part of Artificial Intelligence. Hence, as a starting point for building a robust prediction engine, this research focuses on building a data enhancement engine that would link the plan and model automatically, enabling seamless correlation of the data in the two silos and making it semantic. An engine has been developed as part of this research, and it performs well with respect to the existing data set. Several suggestions to improve the engine have been given as part of the future scope.