Summary & ConclusionsIn the traditional world, field reliability data for the Heating, Ventilation, and Air Conditioning (HVAC) industry, and many other industries’ equipment, provide lagging indicators due to the inherent delays in data acquisition. Warranty databases were the commonly used resources for analyzing field failure percentages and predictions of warranty costs. Warranties takes care of the customer after the damage has happened. Therefore, the warranty provides a lagging unreliability indicator and the failure has a negative impact on the brand name associated with the equipment. In addition, root cause analysis is harder on failed systems due to the catastrophic effect on the system, as well as on the interfaced systems, followed by possible destruction of evidence.Today’s real-time data acquisition systems provide not only real-time failure data but also system degradation and early warning before real failures occur, thus serving as leading indicators. These data can be used to improve systems before failures occur. In addition, improvements can be incorporated into future systems before releasing to the customers. These concepts are discussed in this paper with examples.The survival probability of a system and/or its components is defined as the expected design lifetime. It is a crucial component within the reliability definition. While the system-level design life is commonly in terms of years, the components’ lifetimes are derived in terms of functional usage duty cycles (FUDC) of such subsystems and components. A classic example is the ball bearing lifetime defined in its number of revolutions across 20 years for the 90th percentile customer usage. With the FUDC information, accelerated life testing can be accurately planned. Many industries struggle to define the subsystem and component design life in terms of FUDCs. This leads to ambiguous reliability requirements provided to design teams and critical part suppliers. Big data is of great help to fulfill this need. This paper discusses the concepts of defining the design life in FUDC using big data.Most consumer equipment goes through seasonal patterns due to annual climatic conditions. HVAC equipment provides classic examples of such seasonal usages. Dynamic energy modeling can be effectively used to model seasonal demands and behaviors of HVAC equipment. The big data acquired from cloud-based intelligent HVAC systems and dynamic energy modeling can be developed to have symbiotic relationships. For example, ideal application conditions are not practically available from big data. However, acquired big data from an actual building data system can be used to validate and fine-tune the dynamic energy modeling for that application. Thereafter, energy models can be used to create data for the ideal application conditions. These data can be used to define the design life of subsystems and components in FUDCs. These concepts are discussed using examples in HVAC applications.In summary, this paper presents how big data can be useful to improve and influence leading indicators and improve design for reliability with accurately defined design lifetimes in terms of FUDCs for accelerated life testing. In addition, this paper discusses how dynamic energy modeling can be beneficial in designing for reliability.