Construction planning with machine learning

Paper in proceeding, 2019

Over the next years, it is expected that machine learning will be widely implemented within fields in the construction context, such as construction planning. As construction projects tend to be influenced by interrelated issues resulting in cost and/or time overruns and lower performance, it has been continuously attempted to develop predictive planning methods and tools, in order to mitigate such issues. This study aims at investigating possible applications of machine learning for construction planning, noting their impact on project performance, and finally commenting critically on the issues of responsibility in action-taking, accountability in decisionmaking, and the still crucial need for human reasoning. Methodologically, a literature review on machine learning applications in construction project planning is carried out, and then two particular implementation cases are selected for a more in-depth analysis. The first case draws on a productivity survey of construction projects in Sweden, where the relative data is analysed to find the most influential factors behind project performance; then, statistical correlation is used to find the features that are strongly correlated with four performance indicators (cost variance, time variance, and client- and contractor satisfaction), and a supervised machine learning analysis is done to develop a model for predicting project cost, time and satisfaction. The second case elaborates on the appraisal of constructability of civil engineering projects
through technical project risk analysis; the model utilizes both unsupervised machine learning for the understanding and pre-processing of data, and supervised machine learning for the development of the predictive system. Following the above analysis, it is argued that there is a need for human reasoning in construction planning, even more so after the introduction of machine learning. It is not enough to include human aspects in the machine learning modelling; it is also crucial to strengthen qualified reasoning in the decision-making for construction project planning and being responsible in action-taking and accountable in decision-making.