Once Upon a Time in the Future: Understanding the Risks & Uncertainties of Artificial Intelligence (AI) Development In Complex MEGA Construction Project Environments

Recently, significant construction projects have been announced for extensive AI development, necessitating vast amounts of construction resources. These MEGA projects will require substantial quantities of steel, cement, lumber, electrical equipment, and miles of wiring. What potential challenges could arise during the delivery process? Where will this increased demand find reliable and stocked suppliers? Beyond the estimated material quantities needed for these identified projects, there will also be a tremendous need for utility, housing, and transportation infrastructure to support the construction of these MEGA projects.

When examining the common challenges faced by the project team and estimators, we identify three main categories of risk and understanding that guide the estimates for quantities and costs needed to complete the work:

1. **Known-Knowns**: This category includes the most straightforward estimating challenges, such as the tonnage of steel, cubic yards of concrete, volume of lumber, cost of transformers, amount of switchgear, and miles of wiring required. Develops plans and specifications into quantities and Costs.

2. **Known-Unknowns**: These include uncertainties regarding the location and availability of the labor force, utility expansions needed to support power, water, sewage requirements, as well as permitting and environmental issues. Additionally, there is the competition from other MEGA projects that may also be vying for the same critical materials and labor resources. This is experienced based and is percentages of identified hard costs.

3. **Unknown-Unknowns**: This most challenging category, is often referred to as Force Majeure events, and poses significant difficulties in estimation and management for project success. This is the realm of predicting future events and probability risk analysis. 

The following three illustrative examples are based on our consulting experience and provide some context for consideration of Unknown – Unknowns:

   – A large refinery in the Middle East had just begun construction when Japan experienced the Kobe earthquake. This disaster caused a sudden shortage in the global steel supply as Japan switched from being a major exporter to a net importer. Consequently, construction on the refinery was severely hindered, as design work had been completed without any procurement yet taking place. Recovery from the steel disruption required redesigning key structures and sourcing alternative steel dimensions that did not meet the original design specifications. This unforeseen impact could not be accounted for within the original budget or timeline. Recovery required extraordinary efforts to accomplish.

   – A Multi-Fuel power plant in Asia was close to completion when delays in the transmission system due to poor planning, material shortages, and construction holdups halted its operation at commissioning. The plant remained unused for years while the power grid was extended and  expanded. The expected revenue to cover construction costs did not materialize, rendering the project uneconomical.

   – A high-rise condominium project that seemed poised for early and lucrative occupancy faced issues after residents moved in, citing persistent ammonia odors in their units. Investigations initially failed to find a source for the smell until forensic analysis revealed that changes to the environmental control system (Selective Catalytic Reduction) from a power plant that was providing fly ash additives to a concrete batch plant had contaminated the concrete. Ammonia carryover in the concrete was evolving during curing resulting in elevated ammonia levels in the building. Compounding the problem, the building’s energy-efficient ventilation system inadvertently intensified the issue by recirculating the ammonia. This was an instance of a true Unknown-Unknown that arose from unique unforeseen environmental conditions.

As teams prepare for the impending influx of AI projects, they must take into account all three levels of estimating and environmental challenges. Once projects commence, Construction Managers must remain vigilant for “Black Swan”^[1] events beyond their control that could significantly impact the project’s success. Employing the best practices from Systems Thinking^[2] and Complexity Theory^[3] can provide project teams with the tools needed to detect, respond to, and navigate the Unknown-Unknowns that may arise to challenging the project.

Click here to read more about Systems Thinking and Complexity Theory.

---

1. The Black Swan – Nassim Nicholas Taleb, The Impact of the Highly Improbable.  Random House ISBN 978-1-4000-6351-

2.  Ackoff Center Weblog: SYSTEMS THINKER Russell Ackoff

3. Complexity: A Transient Condition Precedent to Project Failure; Understanding and Surviving Project Complexity: MDC Advisor  – www.mdcsystems.com