By Stephanie C.
Technology has changed drastically in the past twenty years and continues to grow today. Many industries are using up-to-date technology in order to get ahead in the game. In particular, the construction industry has been using a method of data analysis, known as machine learning, to increase productivity and safety, reduce costs, and much more. “Machine learning is a subset of artificial intelligence” that allows computers to learn from previous data without intervention from humans, and it does this by creating algorithms. These algorithms are grouped into two main categories: Supervised learning and unsupervised learning (“Machine Learning,” n.d.). Supervised machine-learning algorithms require the data to be labeled in order for decisions to be made, while unsupervised machine-learning only require the raw data and nothing more. The construction industry can greatly benefit from machine learning, but it also faces challenges along the way.
To begin with, machine learning can benefit the construction industry in many ways. First, the techniques used in machine learning “can be used to optimize development and quality control testing of products and design structures” (“AI and Robotics”, 2016). Machine learning can help the construction industry become more efficient and effective with the given resources. In addition, machine learning benefits the construction industry by giving “fewer errors and omissions, safer jobsites, improved workflows, and more on-time completions” (“AI and Robotics”, 2016). Machine learning allows engineers to simulate construction, such as a building or bridge, which can result in fewer construction errors and reduce costs.
As for tools and software used in machine learning, there are several that businesses can choose from. Some of the common tools and software used include WEKA, R Platform, RapidMiner, and Orange. WEKA and R are examples of machine learning platforms that “provides capabilities to complete a machine learning project from beginning to end” (Brownlee, 2015). WEKA software uses the Java language, while R software uses C, Fortran, and R. RapidMiner and Orange are machine-learning tools that include graphical interfaces, which allow for more focus on aspects such as visualization (Brownlee, 2015). In construction, both types of these platforms and tools are useful, especially the visualization aspect.
The construction industry also has many applications of machine learning. For example, the supervised machine learning technique, logistic regression, can be used “to predict the pros and cons of the Construction Project and allowing the promoters to take any corrective action before damage occurs” (Viswanathan, 2015). By allowing the promoters to avoid any potential damages to the construction project, the process will become more efficient and save costs. Another example is that machine learning is used to predict the amount of damage an earthquake would cause to a building. Knowing the building’s features and the ground motion, an unsupervised machine-learning technique, k-means clustering, can be used to categorize ground motion based on the frequency (Amory, Ferguson, n.d.).
Besides the benefits of machine learning, there are some challenges and limitations that exist. One challenge is that, “particularly in the construction industry, is jobsites are very different from the majority of work places in that most of the work takes place outside in highly unstructured environments” (“AI and Robotics”, 2016). A lot of the machine-learning applications that are available are used in industries with similar operations and indoor environments (“AI and Robotics”, 2016). It is also important to note that all machine-learning methods are imprecise. They do not have one hundred percent precision; machine learning does involve guessing and trial-and-error. In construction, trial-and-error can result in increased costs, just like with any other industry.
In conclusion, the construction industry can gain many benefits from using machine learning with some challenges and limitations. However, despite the challenges construction has with machine learning, the amount of benefits clearly overrides them. The construction industry can definitely be improved with machine learning, especially in terms of efficiency, cutting costs and safety. Machine learning has the ability to make major improvements in construction, as well as positively change human interaction with technology.
AI and Robotics: The Future of Construction. (2016, October 5). Retrieved January 19, 2017, from http://www.conexpoconagg.com/news/october-2016/ai-and-robotics-the-future-of-construction/
Amory, M., & Ferguson, M. (n.d.). Earthquake-Induced Structural Damage Classifier. Retrieved January 19, 2017, from http://cs229.stanford.edu/proj2015/343_poster.pdf
Brownlee, J. (2016, June 06). Machine Learning Tools. Retrieved January 19, 2017, from http://machinelearningmastery.com/machine-learning-tools/
Machine Learning: What it is and why it matters. (n.d.). Retrieved January 19, 2017, from http://www.sas.com/it_it/insights/analytics/machine-learning.html
Viswanathan, Balaje. (2015, April 7). Aligning IOT with Construction Industry. Retrieved January 19, 2017, from https://www.linkedin.com/pulse/aligning-iot-construction-industry-baalaje-ms-viswanathan