The principle message underlying my presentation is about the need to revisit inherited construction engineering practices, especially those related to construction planning and control to achieve lean construction. I will discuss why current practices are deficient, and what we could currently do to correct them, and how artificial intelligence is the way forward.
It is fortunate that lean construction scheduling is increasingly supported by modern tools and methods, evolving in contemporary to advances in Building Information Modelling (BIM). Modular construction, 3D printing, or prefabrication in general could be utilised today more than ever to enhance construction efficiency. Nonetheless, despite these advances, the construction industry still retains a stagnant position to transit from traditional planning techniques, of which most of its roots are traced back to the 1950’s.
The Critical Path Method, for example, was first tested in 1958 on a chemical plant construction. The Work Breakdown Structure was first published in Frederick Taylor’s famous book “Principles of Scientific Management” in 1911. We repetitively see how significant are the divergences between our planned schedules and actual construction, but rarely ask ourselves whether these techniques are appropriate to plan complex construction projects.
I understand that our audience are from different backgrounds. Last year at my 4D BIM speech at The Big 5, I was impressed by the diversity of audience who attended the session. We had attendees from all over the world, and coming from different perspectives, like project clients, designers, site engineers, project managers, students, educators. Lean construction is all about increasing efficiency and removing unnecessary cost. This should be everyone’s obsession.
Considering John Egan’s estimate in his article “Rethinking Construction” that 30% of construction is rework, and that 40% of the manpower used on construction sites are wasted, one could realise how important is to go “lean” from a socio-economical angle. Today we have no choice but to think outside the box and investigate what technology has to offer for reforming our industry.
Research that we conduct at UAE university attempts to approach lean construction from an artificial intelligence perspective. We are developing an inclusive algorithm that we refer to as “self-recovering schedules”. The algorithm utilises machine learning methods such as Neural Networks and Bayesian Belief Networks to train computers about what drives labour productivity based on daily environmental, ergonomic, physiological and mental conditions effecting labourers.
The algorithm updates construction schedules to more realistic ones and hence alerts construction planners about foreseen schedule overruns to take proactive actions in a timely manner. It is very challenging to make computers predict daily productivity given the tremendous number of factors to be considered. Through this interdisciplinary research, we are interested in studying every detail, such as workers respiration rates, blood pressure, glucose levels, etc. Those are measured through a workers gate, a single platform encompassing numerous electrical sensors that reads labour ID with other readings relevant to labour health and mental statuses. I am keen to present the progress of this work beside other relevant lean construction research projects conducted by myself and other researchers.
Dr. Hamad Al Jassmi is an Assistant Professor of Construction Engineering & Management and the Assistant Dean for Research and Graduate Studies at the College of Engineering – UAE University.
He is the first author of several research articles, which have appeared in leading international journals. His research interests include Lean Construction; Artificial Intelligence in Construction Management; and Building Information Modelling. He is particularly passionate about educating future engineers on utilising state-of-the-art 4D and 5D methods and technologies to achieve a longly aspired industry-wide application of integrated lean construction.