Data-driven construction management

Imagine a vision where every decision in a construction project is backed by data, every risk is anticipated, and every stakeholder is aligned. The construction industry has been hurrying to catch up with the technological advancements, software developments, and data that have emerged over the last two decades. Innovations such as drones, artificial intelligence, augmented reality, and many other promising technologies are the focus of research and development projects aimed at integrating them into this massive sector. However, we are still far from seeing these technologies consistently prove their value to construction projects in general. Amid all these breakthrough attempts and the surrounding noise, one aspect I find overlooked in the construction industry is the use of data science. While the industry has been bombarded with the notion of digitalization, there has been little attention paid to the digitization of a construction project environment.

According to the Project Management Institute (PMI), a typical project comprises 10 knowledge areas. For construction projects though, PMI recommends considering five additional knowledge areas specific to the construction industry. While all 15 knowledge areas need to be accounted for and managed throughout the project lifecycle, most discussions and project performance reviews tend to be highly subjective and based on personal interpretations and judgments. During a weekly coordination meeting, it is very common to hear someone express a feeling that the project will be delayed, while another person claims the opposite—both without any sound analysis or quantitative information. In most cases, the basis of the discussion is not solid data but the degree of optimism or pessimism of the claimers.

Benefits of Data-Driven Environment

So, why is it important to transform construction projects into data-driven projects? While there are many different reasons and potential answers to this question from various perspectives, the most fundamental reason lies in the nature of construction projects. These projects involve a large number of diverse stakeholders who interact and impact the project at different levels through different interests and aims. In some cases, these interests may even compete with one another. Additionally, the intensity of involvement and expertise within this stakeholder group varies greatly. All these aspects add to the project's complexity and create a challenging environment for effective communication and analysis across the project stakeholders.

Stemming from the competing interests of various stakeholders, discussions and claims can be highly subjective and speculative, despite the technical nature of the project environment. This situation resembles a group of people speaking different languages, with little interest in reaching a consensus based on facts, but rather in convincing others to align with their individual goals. In this chaotic context, data plays a crucial role in settling the dust and clearing the fog. While statistics are not entirely free from manipulation or misinterpretation, when introduced into a discussion, data can transform the subjective project environment into an objective one. More importantly, it can focus efforts on what truly matters and needs to be addressed, rather than getting lost in unending and unproductive blame games.

Enabling Data-Driven Construction (Study Case)

Transitioning to a data-driven environment may seem daunting, but it doesn't have to be. Data collection can often be associated with sophisticated tools and methods to gather information from different sources. However, before investing in such tools, the first step in the transition to data-driven management can be the simple act of digitizing ongoing activities that can be tied to specific project goals. For instance, due to the hazardous site conditions and labor intensity involved, one of the important project goals is to deliver the project while ensuring the safety and well-being of the workforce and the environment. When considering this goal, it is clear that the most successful outcome would be maintaining a zero-accident record throughout the project.

When thinking about an accident scenario in reverse, there must have been a dangerous condition present on site. This condition would need to remain long enough and unattended to lead to an accident. To assess project performance in this regard, data that is leading or correlated with accidents needs to be collected. Typically, site activities are inspected from a Health, Safety, and Environment standpoint, and warnings of unsafe conditions are given to contractors, who are then expected to remediate such conditions immediately. In a hypothetical ideal scenario, the work on site is expected to proceed while meeting HSE requirements at all times, meaning no warnings should be necessary. By digitizing the content of inspection reports periodically, data on the number of non-compliances, negligence, remediations, and near-miss incidents can be represented quantitatively in addition to the qualitative information. Soon, a set of data points would be obtained to analyze HSE performance and gain insights regarding future risks and performance.

To illustrate this concept further, below is a real-world example from a construction project. The project was a greenfield building initiative where I represented the client, who was both the project investor and operator. Other primary stakeholders included a project management consultant supervising site activities and a general contractor responsible for delivering the project under a fixed-price turnkey contract.

Among the various focus areas of the project, safety was a key priority. This was tracked through several performance indicators, including the number of nonconforming work items and the number of remedied work items. The chart below clearly shows a strong correlation between negligence in remediation and accidents on site. During the weeks when safety issues remained unresolved (yellow bars), accidents occurred (red bars). The trend line for safety warnings (blue line) indicates that the contractor did not improve its safety performance throughout the project. It is also evident that while the project faced safety issues consistently from the very beginning, the contractor took timely actions during the first 30 weeks. However, in the subsequent weeks, specifically between the 30th and 50th weeks, the contractor failed to resolve safety issues in a timely manner.

If the above chart is analyzed alongside the project schedule, a possible explanation for the degradation in safety performance could be that the contractor was experiencing delays on site and attempting to expedite progress while compromising safety. Alternatively, they may have been missing key safety personnel, which could be identified through an analysis of the organizational histogram charts. The crucial point here is that, in the absence of such data, discussions can become highly subjective and speculative, leading to conflicts and unsubstantiated claims. Regardless of our level of expertise or the accuracy of our assessments, without data, discussions can easily turn personal and offensive.

Let the data speak the truth.

Even though data-driven management in construction projects promises significant potential for improving overall project effectiveness and optimizing resource use toward achieving strategic goals, a primary barrier to transitioning the industry lies in its change-resistant nature and the traditional mindset of professionals. Despite the collaborative nature of construction projects, which involve multiple disciplines, construction management practices are often limited in their interactions with and inspirations from other industries. For instance, although every step in a construction project involves decision-making, the application of scientific decision-making methods is rarely seen. Similarly, despite the presence of many engineers and analytical minds, there is often a lack of recognition for the need for business analysts or data analysts within construction projects. This highlights the necessity for construction industry professionals to develop new skills and enhance their data literacy.

The Future of Data in Construction and Sustainability

In addition to its professionals gaining data skills, construction projects will need to start including new roles and positions focused on data, analysis, and design thinking to implement data-driven management. From a methods perspective, during the data-driven transition, construction projects will increasingly incorporate scientific methods from different fields for decision-making, multivariate data analysis, and system dynamics, each improving different aspects of the project domain. For instance, by utilizing multivariate data analysis on different factors and variables of a project environment, it will be possible to find correlational and causational relationships between different constructs of a project environment, discover their influences, and establish predictive, explanatory, and confirmatory models of projects. If such a breakthrough in the construction industry occurs, we may witness highly accurate and predictable project environments where meeting project goals will be more achievable due to increased clarity and a reduced number of unaccounted risks.

As construction projects embrace data-driven management and integrate scientific methods, it is important to recognize that the human element—emotions, desires, and biases—will still significantly influence project outcomes. At our core, humans are as emotional as we are rational, often justifying our actions through feelings. Additionally, there is a constant struggle between short-term and long-term decisions, each yielding different results in terms of pleasure and pain. When we examine our environment, significant waste of precious resources becomes evident, often driven by human urges or weaknesses. For example, even when someone has a perfectly functional phone, they can find numerous seemingly innocuous reasons to justify purchasing a new one. Whether the motivation is an attempt to showcase social status or not, they might easily claim that the current phone is unreliable or lacks features—such as a few extra screen pixels—that they deem vital for daily use. Interestingly, this same individual might also deliver a passionate speech on sustainability and the world's suffering from the waste of finite resources.

The same can be applied to the products of the built environment, such as buying a third house or building an ultra-luxurious mansion. There are a number of important aspects needed to transition any industry into a sustainable one. For the built environment and many other industries, one of these aspects is the sharing economy, which necessitates a change in personal and social behaviors and choices. A choice can be made to build an expensive summer house in a naturally preserved area and use that building once a year. Alternatively, one could find a time-sharing solution to spend summer holidays in a social setting. It is obvious that the latter is more sustainable than the former, but one can easily find reasons to justify choosing the unsustainable alternative. Similarly, in a corporate space setting, there might be a need to accommodate the expanding headcount of an organization. However, the solution might be increasing space efficiency rather than building more spaces. From a systems perspective, this often comes down to a decision-making trade-off between the comfort of having personal space and the social collaboration of using common spaces.

Next

In all the discussions above, if introduced with clarity and transparency, data can significantly help us create the best and most optimal solutions. In the case of building a summer house, data on feasibility can provide insights into the effectiveness of the investment compared to the time-sharing alternative. Similarly, a space and utilization analysis can enhance space efficiency without necessarily requiring additional resources. Ultimately, it is up to us to decide whether to make choices based solely on emotions and feelings or to consider data-driven insights, regardless of how well they align with our human desires. One thing is certain: without the use of data, sustainability cannot be achieved, as discussions would remain personal and subjective, lacking a common ground for benchmarking or measuring improvement.

This reliance on data is part of a broader context of rapid technological advancements that redefine our understanding and communication. The new millennium started with unprecedented advancements in information technologies, and

data has become the new language of this era.

In this new era, data not only creates a unique and robust communication environment but also helps us understand the world beyond our perception. It can overcome obstacles arising from the subjectivity of human feelings and desires, allowing for clearer insights. Whether we acknowledge it or not, digitalization, sustainability, and data are converging to form a new ecosystem. Contributing to this ecosystem can be as simple as representing and tracking our goals and actions in plain numbers. This straightforward step often leads to further exploration, discovery, and understanding of our systems. Interestingly, the very urge that sometimes holds us back—such as the fear of being left behind—may also drive us to engage in this transformative process.