Time is of the essence when it comes to the role of an expert in a construction dispute. With cases increasingly involving vast amounts of data, AI has become a formidable tool for these experts, significantly improving the efficiency of data analysis.

However, AI tools are not without their limitations. Human oversight is still crucial to ensure both reasonableness and accuracy in legal proceedings. In each instance, trial and error is necessary to validate the desired output. In this article, we explore the role of AI in construction disputes and the potential risks around its use.

Forensic Delay Analysis Considerations

Forensic delay analysis in the construction industry involves a detailed process carried out by experts to determine the causes and extent of delays in a project schedule. This process includes:

1. Data collection and review: Gathering all relevant information related to the project.
2. Source validation: Validating the reliability and accuracy of the available schedules.
3. Delay event identification: Pinpointing specific events that caused delays.
4. Delay event analysis: Examining the nature and impact of these events.
5. Delay quantification: Measuring the duration and extent of the delays.
6. Responsibility assessment: Determining which parties are responsible for each delay.
7. Financial cost calculation: Estimating the financial impact of the delay period.

Traditional methods of forensic delay analysis involve manually reviewing extensive project documentation, a process that is time-consuming by nature. Similarly, identifying specific delay events often requires a painstaking review of schedules and logs, which can take weeks or even months.

AI and natural language processing (NLP) tools optimize this process by automating data extraction and analysis, following the processes and steps established by the expert. For instance, some AI tools can quickly parse large volumes of data using sophisticated queries to identify relevant information and anomalies. These automation tools not only speed up the research process – by streamlining data collection and review– but enhance the expert’s ability to gain a clearer understanding of project issues.

Another technology enabling experts to more effectively review large datasets is machine learning. Traditional methods reliant on static historical data and manual analysis can miss emerging trends or overlook subtle indicators of potential problems. In contrast, some AI tools can use machine learning algorithms to continuously analyze both historical and current project data. These algorithms can detect deviations from the norm and identify (based on the frequency of occurrence of certain events) common delay causes such as supply chain disruptions, weather events or resource constraints.

Additionally, machine learning models can assess the reasonableness of project schedules, comparing them against historical performance data and industry benchmarks. From a forensic delay analysis perspective, these tools can simulate various scenarios to evaluate the impact of different scheduling assumptions and constraints, assisting experts in determining whether proposed schedule updates were realistic. They also help identify potential optimism bias and check if project timelines were achievable and aligned with performance trends.

4D BIM

By integrating time with 3D project models, advancements such as 4D Building Information Modelling (BIM) mark a leap forward from traditional scheduling methods. Unlike static Gantt charts, which often lack insight into how delays impact the overall project, 4D BIM and Augmented Reality software are dynamic tools that allow users to visualize as-built progress over time. This interactive capability helps identify potential scheduling conflicts and impacts of delays, supporting delay management and mitigation.

While the core methodologies of forensic delay analysis – as outlined by industry guidelines such as the Society of Construction Law, Association for the Advancement of Cost Engineering, and American Society of Civil Engineers – will continue to slowly evolve, advancements in the tools and technologies used by experts have the potential to transform the construction dispute landscape. 

With AI, NLP, machine learning and 4D BIM more integrated into forensic delay analysis, experts will be able to deliver opinions and reports quicker, with more evidence and data available to support the experts’ findings. Over time, contractors and owners will find these innovations will also enhance project performance, making delays easier to predict and manage. Consequently, the construction industry can expect greater transparency, fewer disputes, and more efficient project outcomes in the long term.

Quantum Considerations

The common starting point for an expert focuses on the three key quantum principles:

1. What was the actual cost?
2. Was the actual cost reasonable?
3. Can the actual cost be allocated to the various causation factors?

In large, complex disputes, the expert is instructed to consider huge volumes of data in order to substantiate costs incurred. These files – sometimes in excess of 500,000 documents – typically consist of cost system downloads, invoices, purchase orders and other records. Large teams of professionals are normally required to handle, review and analyze this data.

One way to minimize the manual effort of large-scale review is to use AI in data sampling. The technology is already capable of handling huge amounts of data which can aid with the sorting and reconciliation of the data available, as well as identifying issues such as duplication.

In terms of analyzing reasonableness of the costs, this will depend on the experience of the quantum expert. What is interesting however, is the potential to use AI data to compile data analysis and benchmarking of construction costs, such as labour, plant and materials. Predictive modelling is another function AI can provide, forecasting costs based on specific locations, market conditions and other relevant factors. Through optimization algorithms, AI can analyze market trends and economic indicators to provide insights into future cost fluctuations. In the case of dispute avoidance, this information can be used to adjust cost estimates and ensure they remain reasonable over the course of the project.

When analyzing various causation factors, AI can be used to aggregate and analyze large volumes of data from various sources, such as project management software, financial records and communication logs. By examining this data, AI can identify patterns and correlations between cost overruns and specific events or decisions that may have contributed to a dispute.

The evolution of technology, including AI, enhances the output of the quantum expert, allowing them to use their industry knowledge and experience to interpret AI analytics and provide a clearer picture of the root causes of cost-related disputes, and to be more cost effective and efficient.

Risk Considerations

While there are benefits around the use of AI in construction disputes, there are also potential legal and regulatory ramifications that experts should be aware of when leveraging these tools. These may occur as a result of three shortcomings:

1. Overreliance in AI data in an expert’s work product
   AI and machine learning are additional tools in practitioners’ toolbox when evaluating construction disputes. However, these technological advancements should not replace the role of an expert in assessing the delay and quantum events.
2. Errors and omissions in the AI setup and data output
   The proper setup to review data is a laborious process. The automated output must be verified before its consideration. As the scope and magnitude of datasets increase, so does the time to properly debug and troubleshoot until accurate results are achieved.
3. Deviations and gaps in the project data analyzed
   Deviations in the data source are frequent on construction projects facing delay, quantum, or both. It is not uncommon for troubled projects to have data in differing formats or for datasets to be missing altogether. For example, construction schedule updates are often provided in formats other than the native software (.pdf file format in lieu of .xer format).

If the expert is not cautious when performing the required analysis for the dispute, each of these shortcomings can result in the loss of credibility in the specific engagement, the removal of the expert from the current case, as well as introduce potential long-term impacts to the expert’s reputation for misuse of the technology.

The advent of AI is a welcomed addition to the playbooks of construction dispute experts, given its power to increase the efficiency by which voluminous sets of data are analyzed. The available enhancements to the expert should be accepted, but they are not a substitution for expert analysis.

This article was first featured in Mealey's Litigation Report: Construction Defects.