Objective - To improve the operating efficiency of commercial heating, ventilating, and air-conditioning (HVAC) systems by developing automated fault detection and diagnosis (AFDD) tool algorithms as a new measurement science applicable to building performance, while also transferring that measurement science to the commercial marketplace as a practical technology by way of immediate and ongoing collaborations with private–sector partners.
What is the new technical idea?
Based upon its recognized legacy work in fault detection and diagnostics, NIST has established two cooperative research and development agreements for partnering to develop novel AFDD tools able to leverage the untapped capabilities of modern automation and control systems in order to improve the energy performance of buildings. Recent significant advances in the technologies that automate buildings provide the prospect of data–rich performance surveillance, that can be applied on the system-wide scales necessary to optimize overall building performance and achieve the goal of net–zero energy consumption. Building owners have so far found the building automation industry itself to be notably challenged in developing and bringing such tools to market. The new technical idea is to apply a cross–disciplinary mix of technical expertise in artificial intelligence, the thermodynamics and fluid mechanics of the mechanical processes (e.g., HVAC) under surveillance, and the skills to formulate and express that blended expertise in the form of extensible, maintainable software applications able to run on multiple computing platforms and interface with varied data sources to develop the measurement science for the next generation of AFDD tools.
What is the research plan?
NIST will enhance the relevancy and transferability of its proven expertise in AFDD, and in particular, its expertise in applying and automating sophisticated artificial intelligence methods. This will be done by establishing within NIST a new collateral competency at formulating and expressing its AFDD methods via modular, serviceable, object–oriented program (OOP) software components that match, and thus can seamlessly integrate with, the state–of–the–art software design practices used in complementary application products marketed by NIST's private–sector partners. The result will be new AFDD tools that adapt readily to varied computing platforms and data sources, while being extensible to accommodate new system or equipment applications and upgradable as the methods the tools execute are further developed. Through collaborations with key partners in both the public and private sectors, NIST will develop and test, at multiple field locations, prototypes of such AFDD tools, proving "expert system" and other techniques from artificial intelligence can be implemented via practical "real–world" AFDD tools that help improve the efficiency of buildings.
NIST will set out on this path by its project staff following these principal steps:
- Acquire and apply contemporary object–oriented software design skills as needed to recast into a more modern, more serviceable and transferable form the code architecture of "FDD Expert Assistant" AFDD tool (developed under the prior project, "Fault Detection and Diagnostics for Commercial Heating, Ventilating, and Air–Conditioning Systems").
- Develop an improved desktop PC–based rapid development environment for testing, evaluating, and revising AFDD tool algorithms, where the algorithms can be run using the identical code components used to embed the tools in the targeted field hardware. This will be much more efficient than the current development environment that requires complete recoding of promising algorithms before they can be applied in an embedded system.
- Once the above two steps have progressed enough to allow it, further development of AFDD tool analysis algorithms and associated features will proceed under two existing CRADAs with private–sector partners. They will provide valuable feedback and technical assistance on tool usability and interface compatibility, while benefiting from NIST expertise in the tool algorithms.
The development of new algorithms and features will focus on areas identified as important during collaborative testing with the CRADA partners. It is tentatively planned that these will include tool reliability enhancements aimed at robustness to spurious data, prioritizing for the user the economic impact of faults found, and enhancing autonomy by replacing some queries of the user for diagnostic information with direct references to new knowledge base elements.
 Turner, C. and M. Frankel. 2008. "Energy Performance of LEED for New Construction Buildings. U.S. Green Building Council.