Man vs Machine (Learning): The Problem with One-Time Audits

Posted on

According to the World Health Organization (WHO), “Health is a state of complete physical, mental and social well-being and not merely the absence of disease or infirmity.”1 The WHO’s definition of health recognizes that health is more than just the measurement of illness, but should reflect the general well-being of each system that makes up the whole. Buildings are a lot like humans in this respect – they’re large complex systems made up of many smaller systems. In order for a building to truly be healthy, we need to have the right set of analytical tools to understand how a building, with all of its intricacies, performs.

In the 1970s, the cost of oil skyrocketed and forced the United States and many other countries to examine how energy efficiency was measured and improved in the largest industrial sectors – including buildings. Thus, energy audits were born. Energy audits have remained the standard industry tool for evaluating the performance of a building; however, the idea of surveying buildings for the purpose of improving performance has since been expanded to include additional methodologies such as commissioning and retro-commissioning.

While energy audits and commissioning-based approaches can both identify and correct deficiencies that show a measurable impact almost immediately, they often leave many issues undiagnosed. This is not to say that energy audits and retro-commissioning do not have a place in the industry, rather, that they are simply limited in their methodology. As building management system technology has evolved, so has the desire for a more sophisticated way to evaluate, measure and improve building performance.

manvsmachine As Lord Kelvin said, “If you cannot measure it, then you cannot improve it.” The problem with traditional methods of analysis is that the potential of big data sets is largely missed by engineers using spreadsheets and basic tools, as they can only see the tip of the iceberg. To put it simply, big data equals big potential. The problem with traditional methods of analysis is that the capabilities of handling large sets of data quickly exceeds the abilities of engineers – hence, big data can lead to big problems. Building management systems and HVAC equipment are now capable of generating large amounts of data. This data has the potential to provide highly granular insight not just for individual equipment sub-systems, but for building performance as a whole.

In the past 10 years, the idea of continuous commissioning has surfaced to address the limitations of the traditional auditing methods. With the increase of more highly sensed systems and improved data export capabilities, it is now possible that a building can be monitored continuously through the use of software, thus extending the benefits of an energy audit over an indefinite period of time. The main obstacles to widespread continuous commissioning have been the handling – and efficient analysis of – gigabytes of data.

Managing big data is not a problem unique to the building industry. Systems in many other industries are generating large sets of data and need a solution to manage and analyze this data. The term “Internet of Things” has become popular to refer to the networks of interconnected devices that communicate and exchange data. Similar to buildings, efficient analysis methods are required for generating insight into the performance of different devices and the systems they make up.

Over the past five years, startup companies in the private sector have emerged to tackle the challenges of handling big data, performing effective analytics, and measuring results.

Enter Ecorithm. Ecorithm’s first product is a solution that utilizes domain expertise in the building industry to improve the performance of buildings. Ecorithm enables any building, no matter shape or size, to use the power of advanced analytics to continuously identify deep-rooted flaws and to truly understand the health of a building. Ecorithm’s proprietary technology uses various methodologies including spectral, physical and spatial analysis, as well as machine-learning. Using these techniques, massive amounts of data from building management systems can be compressed and analyzed automatically, and a deeper understanding of building science can emerge. Big data needs big software.

Americans spend up to 90% of their time indoors.2 A little alarming, right? With a majority of our lives spent inside, ongoing analysis of building data is extremely important – the interaction that we have with buildings is intrinsic both to our health as well as to the health of our buildings. Ongoing analysis of big data is shaping the way we interact, live, work and socialize; Ecorithm is poised to define that shape.



NOTE: This is Part 2 of a 3-part series describing the state of the industry for each of Ecorithm’s 3 building optimization steps (outlined in this post) and how Ecorithm uniquely overcomes these challenges – and other pervasive issues – to deliver highly accurate intelligence.

1. Identify and Fix Issues
2. Maintaining Health
3. Optimizing for Peak Performance

Keep a look out for the last piece in the series!

Email this to someoneTweet about this on Twitter0Share on Facebook0Share on LinkedIn13