By Ray Diwakar
From the June 2020 Issue
The shift towards higher rates of data digitization and technology utilization has been occurring in recent years, and as the world has grappled with the novel coronavirus pandemic the need is more acute than ever. Across the spectrum, industries are adopting creative ways to solve real problems using new technologies. Construction industry market trends point to an increased use of data, analytics, and artificial intelligence (AI)-based solutions to automate many processes in the construction life cycle, reshaping construction, renovations, facilities management, and maintenance. Additionally, facility executives are looking for new technology applications to reduce construction project costs, renovations, remediations, project planning, improve maintenance scheduling, and to alert them about risks. The efficiencies and benefits from these technologies are immense. They are also entirely dependent upon data.
Data is a critical component of any operational entity. With the advent of newer technologies, such data is growing at an exponential rate. As a result, when most of this data is distributed across multiple platforms, with manual or non-existent workflows and siloed in systems that don’t “talk” to each other, the data can quickly become stale and disjointed.
As facility managers and owners look to improve productivity by becoming more fluent in technology, it is critical that adequate attention is paid to the quality of the data, how it is managed, and how frequently it is refreshed. Having a sound information architecture (IA) strategy across disparate processes, such as managing projects, creating bids, estimating material, and labor costs, etc., is key to realizing the true potential of the great technology shift happening in construction. An essential concept to remember is “IA before AI.” Successful adoption of more efficient and innovative technologies such as AI by various stakeholders—contractors, operators, owners, facilities managers, and service providers—is entirely dependent upon automating manual workflows, reducing siloed systems, and enriching and refreshing stale data. Antiquated systems, manual processes, inaccurate and incomplete data can ruin a well-intentioned AI or an analytics strategy. This is why information architecture (IA) must come first.
Overcoming data silos and disparate information systems is achievable through standardization, leveraging up-to-date information, and investing in data integrations. A high level of focus on data and information strategy helps ensure that investments in advancements like AI, IoT, Building Information Modeling (BIM), and enhanced project management software have a positive return on investment.
A successful data strategy requires better architecture, and augmenting internal data with enriched, continuously updated information in order to deliver results with AI. Additionally, any artificial intelligence involved in making decisions generally made by humans will need to understand construction language or construction vocabulary.
For an AI initiative to be successful, an AI system will need to be trained on common typos, frequently used sentence structures, abbreviations, acronyms, differences in terms, trade specific terms, and unique naming and phrases for materials, equipment, and labor. For example, the AI will need to learn that gypsum board, dry wall, and sheetrock are material names that are used interchangeably. A fully functioning AI system must have an adequate set of training data or historical records that it can learn from and provide automation. Such historical records, if they exist, have to be transformed in the aforementioned data and information strategy initiatives before letting an AI system train. Without first investing in ensuring that data is structured, clean, and current, any AI system will fail to produce benefits in a timely manner. In addition, not only is it important to provide solid training data to an AI system, it is paramount that there is a steady feed of similarly rich data in order for it to learn and develop continuously.
It’s equally important to understand that data-driven AI solutions are tools used by facilities stakeholders, not replacements for human knowledge and intuition. Processes throughout the construction life cycle, such as estimating construction costs and creating proposals, are repetitive, lengthy, time consuming, and prone to errors due to the number of steps involved. With AI-based machine learning techniques, preconstruction and facilities planning workflows can be streamlined. Combined with a successful data strategy, complete with data standardization, systems integrations, and up-to-date information, AI can improve facilities management workflows and processes.
Artificial intelligence can deliver on the promise of the future through a combination of deep learning, natural language processing, data science, construction vocabulary, and “rich” data. But a well-designed data strategy must be the first step. Those who can begin to understand and utilize the benefits of AI and other advanced technologies will be rewarded with short- and long-term benefits as the first to embrace facilities of the future.
Diwakar is vice president of innovation at Gordian, a leader in facility and construction cost data, software, and services for all phases of the building life cycle. He guides initiatives focusing on artificial intelligence and advanced analytics to solve business problems and generate new value for internal and external stakeholders. Diwakar has more than 20 years of experience designing and developing software in artificial intelligence/machine learning, big data, semantic data management and encryption. He earned his Bachelor of Science in Civil Engineering from Osmania University as well as a Master’s in Architecture from Washington State University.
Share your thoughts in the Comments section below, or send an e-mail to the Editor at [email protected]
Want to learn more about technology and facility management?
Check out more technology and facility management news in previous Facility Executive Tech & FM Columns.