Coastal Construction, based in Miami, adopted Togal.AI, a software that employs artificial intelligence (A.I.) to hone estimates. This has saved Coastal Construction about 10,000 hours of labor. With the software, the company can be more competitive and redirect labor to other tasks to breed efficiency, save costs and be more profitable.

A.I. goes mainstream

A.I. is a hot topic and is making huge inroads into industrial sectors. As a mathematical tool, it sorts large volumes of data in a way that allows “learning” to occur with a capacity that exceeds the capability of one human brain. 

Across every sector, there’s an all-out sprint to harness A.I.’s power to build efficient business practices. The construction industry is no exception, and helpful data such as weather conditions, material costs, usage, labor expenditures, schedule data, change order data and more can be collected.

A.I. is used in estimating software and other data science applications in construction in a way that is boosting the bottom line of projects that have been somewhat unpredictable. A.I. can be used to help preserve resources, make better estimates and produce more fruitful project results.

Clark Construction Group LLC, Bethesda, Md., uses Destini Estimator software with A.I. as a management tool in its projects throughout the United States. Data smart computing with A.I. helps Clark improve construction budgeting and planning.

The rise of large language models

A new trend in data science is the emergence of large language models (LLM) for use in decision science to help accelerate A.I. applications within industrial sectors.

These models harvest data using tools such as OpenAI’s ChatGPT and others. A.I. accelerators process and transform large volumes of data into useful information to drive operational decisions.

Structured and unstructured data

LLMs enable the processing of unstructured data such as comments written in a quality assurance report or descriptions of conditions found. Structured data refers to more discrete data such as temperatures, pressures and other quantitative readings, or perhaps survey data selected from a finite number of response choices.

Unstructured data is not as discrete and easy to compute as structured data. It may be text from an inspector’s report, or descriptions written from one person’s perspective. 

With A.I. accelerators, unstructured data may be processed using “natural language processing” in which the A.I. learns the language and can translate it into objective information, which may then be used to drive decisions.

Construction applications

What’s new about all of this is A.I.’s arrival in the production environment of construction sites. Such environments always have many moving parts—with many people engaged in concurrent tasks, all synthesized by data that helps characterize conditions and actions that can be used when tasks are repeated or implemented in the future.

With an LLM, specific information may be extrapolated from reams of data that can tee up useful material for production planners, estimators and other managers.

By employing an LLM with an A.I. engine of sorts, teams may discover the best cable routing path, how much material to order, how many people to deploy to most efficiently accomplish a job and what sequence of work may be most fitting. It now also allows unstructured and structured data.

“A.I. penetration has been low, or we should probably say it’s starting only now,” said Shreesha Ramdas, CEO and co-founder of Lumber, Saratoga, Calif., a workforce management platform powered by a purpose­-built LLM for small and medium construction contractors. 

A.I. is everywhere, improving estimating, budgeting and planning. There’s more to come as A.I. applications are refined and larger data models are used. For the electrical contractor, the applications may potentially be a game changer in the use of data to drive more efficient operations and work practices.

To learn more about A.I. in electrical construction, check out “An A.I. Revolution” in the October 2023 issue of ELECTRICAL CONTRACTOR.

stock.adobe.com / AndSus