AI for Building Systems Innovation (AIBSI) Program

Summary

Buildings account for 37% of the United States’ energy use, more than the transportation or industrial sectors [1]. Over 80% of a building's life cycle energy use is associated with operating the building, rather than the materials and energy used for construction [2]. “Building systems almost never achieve their design efficiencies at any time during building operation and their performance typically degrades over time” [3]. Building control companies, equipment and system manufacturers, energy providers, utilities, and design engineers are under increasing pressure to improve performance and reduce costs by developing building systems that integrate more and more building services, including system reliability and cybersecurity, energy management, fire and security, vertical transportation, optimal control, the real time purchase of electricity, and the aggregation of the building stock. Measurement science is lacking to enable these systems to have the intelligence to communicate, interact, share information, make decisions, detect and respond to equipment failures and abnormal conditions, and operate in a synergistic and reliable manner. Specific needs include standard data models, communication protocols, user interface standards, cybersecurity procedures, testing tools, and performance metrics. Overcoming these barriers is critical if building systems are to meet these operational needs and if the U.S. is to obtain a significant share of the developing worldwide market for such systems. This program will provide the measurement science to realize cost-effective building operation through integrated sensors and building control systems with distributed AI-enabled intelligence that can optimize building system performance, detect and respond to equipment failures and sub-optimal performance, and enable integration of building systems with electric grid technologies. Additional benefits include the potential to automate tasks to alleviate worker shortage, and the means to provide solutions for large building complexes and community-scale needs.

Description

Objective
To develop and deploy advances in measurement science that will improve building operations to ensure cyber-secure facilities, system reliability, occupant comfort/safety and lower operating costs through AI-enabled building systems.

What is the problem?
Buildings are complex systems of multiple interacting subsystems, and these interactions are often not well understood. Most commercial buildings are “one-off” designs with unique operating needs. Furthermore, the sector lacks simulation tools that can realistically capture all of the necessary details of complex interacting building systems. While all buildings need to address these challenges, the building industry is very sensitive to the first cost of new technologies, and not all projects can afford the most sophisticated solutions. Performance goals such as reducing energy costs and improving indoor air quality and thermal comfort can conflict.

An integrated portfolio of measurement science capabilities is needed to support innovation in the design and manufacturing of individual components and systems and to capture the complexities and interactions in real building systems. Each individual measurement capability presents technical challenges, and the overall goal of significantly improved building system performance can only be achieved by applying an integrated portfolio of such measurement science capabilities. This portfolio must include solutions for projects having a range of budgets.

Because a mismatch exists between who invests (manufacturers, building owners) and who benefits (public and other end users), broad stakeholder involvement is necessary to overcome the initial barrier of developing the measurement science. EL is well-positioned to engage these stakeholders by leveraging its strong ties to industry partners, academia, and standards organizations. EL has the necessary technical expertise and an international reputation for excellence in the technical areas relevant to cybernetic building systems owing to over three decades of technical work and collaboration. EL staff holds leadership positions on the key U.S. and international committees to translate program results into practical solutions that include industry-consensus standards and guidelines.

What is the new technical idea?
The new idea is to leverage AI-enabled building systems to address the measurement science needs of cybernetic building systems in a holistic, integrated manner that considers complex system interactions and their impact on system reliability and energy costs, while maintaining occupant comfort, safety, and system cybersecurity. Measurement science is needed that will:

• Lead to enhancements in communication protocol standards that enable the practical use of integrated HVAC, lighting, security, vertical transport, energy management, and emergency response systems to achieve increased comfort, safety, and energy efficiency;

• Enable the use of artificial intelligence (AI), machine learning, and statistical analysis to detect and diagnose unwanted operating conditions that waste significant amounts of energy and money during daily operation for both residential and commercial building applications; and

• Provide the building services cybersecurity tools that are needed by building owners and operators to understand threats, risks, and countermeasures. and a governance approach to ensure cyber-secure facilities.

Manufacturers of building control products and control systems will benefit from this work as they plan and develop new products and services. Standards development organizations will benefit from the technical assistance we provide to improve existing standards and develop new ones that fill technical gaps. Electric energy providers will benefit from the assistance in transforming the industry to providing AI-enabled microservices. Finally, building owners and occupants will benefit from improved building operation and reduced operating costs.

What is the research plan?
The research plan consists of a portfolio of interrelated projects that focus on key areas of measurement science that are necessary to achieve successful integration of artificial intelligence to building reliability and performance, as shown in the following figure. The projects leverage and enhance enabling technologies for the application of artificial intelligence in building systems. Cybersecurity is a fundamental requirement to ensure secure buildings and semantic interoperability enables the translation of data into information for practical applications. New research focuses on developing technologies to improve the practical application of digital twins and a suite of microservices for a variety of building systems. Current research addresses advanced controls, reliability of mechanical systems, and grid reliability. Future work, indicated in the figure by dashed lines, may include building-related aspects for data centers and quantum computing. These collective research efforts provide a comprehensive approach that will lead to new industry standards and practices, resulting in a radical market transformation in building system design and operation.

Credit: NIST

The development of the measurement science requires laboratory testbeds capable of whole building emulation of normal operation and a variety of conditions suitable for evaluating the needs and performance of building systems in identifying and responding to equipment failures and abnormal conditions.

Two unique laboratories play this critical role in the research program. One is the Intelligent Building Agents Laboratory (IBAL). The IBAL is a complex "building in a laboratory" that includes mechanical equipment typically found in a small office building, as well as equipment to deliver conditions and thermal loads consistent with this application. The IBAL supports research in the use of artificial intelligence to optimize the building-scale performance of mechanical systems. The second laboratory is the Virtual Cybernetic Building Testbed (VCBT). This realistic, whole-building emulator combines actual building control equipment with simulated building systems and weather. The VCBT is used by projects in the program to conduct research under controlled conditions that cannot be accomplished in actual buildings. These two laboratories can also work together – commercial control programs in the VCBT can operate equipment in the IBAL

This program also provides technical support for the ongoing development of key enabling standards that create the communication infrastructure used in the VCBT and upon which embedded intelligent systems will be built. EL’s past work has led to international adoption and commercialization of BACnet, one of the most widely used and successful standards in ASHRAE history. EL will continue to work with industry partners to enhance BACnet capabilities in ways that eliminate barriers to extending BACnet beyond HVAC applications and enable BACnet systems to provide operational data to enterprise management tools.

---

References

[1] DOE Energy Information Administration, https://www.eia.gov/

[2] International Energy Agency, Buildings. https://www.iea.org/energy-system/buildings

[3] Federal Research and Development Agenda for Net-Zero Energy, High-Performance Green Buildings, NSTC Report, October, 2008