Is Big Data the Path to Big Savings.
In most cases, outliers are disregarded as something outside the normal range and excluded from
an analytical and management perspective. W. Edwards Deming, the grandfather of Six Sigma, took
a different approach; he said outlier identification was a critical tool in improving a process,
any process. In the 70’s and 80’s, the Japanese auto industry drastically improved product quality
and lowered costs, due in large part to following Dr. Deming’s philosophy, while the U.S. auto
industry largely ignored Dr. Deming. We all know how that worked out.
Fortunately, U.S. manufacturers of heating, ventilation and air conditioning (HVAC) did not suffer
the same fate. However, building owners and managers are making the same mistake when it comes
to using out-of-date processes to manage and maintain HVAC systems once installed. According to
U.S. Department of Energy studies, more than half of all HVAC systems are operating below their
design specifications because --the process. If we apply Deming’s philosophy to the management
and operations of HVAC systems, will we find the Holy Grail to the nation’s energy pains?
Problem definition is more than half the solution.
To define the problem in HVAC operations and maintenance, we need to identify the “outliers”-- the
anomalies that plague otherwise high performing equipment and hardware, using available data.
How much data do you think is available to the HVAC service tech on our homes and small and mid-sized
commercial buildings? -- practically zero. Oh, they can measure various parameters while they are
on site, but they don’t have detailed historical data to compare to current operating conditions much
less how well each HVAC is operating as compared to other HVAC systems. There is no way to find an
outlier, because there is insufficient data to establish a detailed baseline on all aspects of it’s
performance. Although the auto industry took a beating in the 70’s & 80’s, they learned! Not only
have they applied the teaching of Dr. Deming to improve product quality and price, but have loaded
cars with sensors and diagnostics to collect real-time data to keep them running better longer.
"Without data, you are just another person with an opinion."
W. Edwards Deming
So the obvious question now is why don’t we have better HVAC operating data? An HVAC system in
any building is a complicated process that very few understand well. It is difficult to decide
what data to collect, how to collect it and what sort of analysis should be done after data has
been collected. It’s also very expensive to put traditional hardware and software in place that
has to be engineered, designed and installed, from scratch in every building. Sub-meters and
utility bill analysis help, but they miss most of the important data especially when it comes
to HVAC package roof top units (RTU’s). Without a process to capture detailed performance data,
it is virtually impossible to keep the outliers from reducing energy efficiency, performance
and useful life.
Additional Cyber security risk?!
Collecting data also creates an additional cyber risk as you open up the traditional IT networks
and firewalls that contain accounting data, email traffic and customer names and credit card
numbers. What company isn’t spending lots of money to protect their IT networks from cyber-attack?
It certainly is possible to collect data and create a separate virtual or physical network, but
in general, more cyber security means more money.
Here comes the revolution.
As energy decisions move from the boiler room to the boardroom and LED lights put HVAC systems in
the spotlight as the single largest energy consumer in most buildings, it’s time for a revolution
in the HVAC maintenance industry. What will this revolution look like? Well let’s see, almost
all revolutions these days look like this—more technology, more data, web-based, user friendly,
time saver that costs less and does more. Did that capture everything? Oh, two more things, cyber
security concerns have to go away and it has to be “commerce-enabled”.
Resistance is futile
The state of California requires that every RTU manufacturer offers an option to collect data from
the new units. While the current industry paradigm is to invest in more IP addressable thermostats
and expensive building energy management systems, these investments do not address 75% of all HVAC
problems that require a “truck roll” because they do not adequately address two of the most common
problems with HVAC systems—human error and mechanical failure. In addition, control systems have
done a poor job of monitoring themselves. Fortunately, with the advancements in hardware, sensors,
communications and cloud-based servers, purpose-built “one-design fits all” systems are cheap enough
to be installed on every HVAC system and identify all three common problems--mechanical failure,
controls error and human error. As the data begins to reveal these truths about the process, there
will be a landslide of “Deming-esque” activity, driven by the market’s desire to improve the process,
and ultimately reap the rewards of such a revolution. Those who embrace Deming’s philosophy will
likely ride the wave of the revolution and capture market share, while those who don’t will be
wondering why they lost their job—like many former American auto workers.
More to come
Stay tuned for follow-on blogs that will go into detail regarding some of the topics mentioned above including:
- How to collect Big Data inexpensively without cyber-security risks
- Improve HVAC performance, maintenance and repairs
- Digest the financial implications of DOE studies detailing the high costs of doing nothing
- Chart a clear path forward to more comfort, health, productivity lower energy and emissions resulting in higher NOI and shareholder value.