|
Looking Ahead, Looking Behind
This
is the first of a series of articles looking back on the roots of the
IoT, as expressed in Automated Buildings, and looking for the theme we
need to consider for success going forward. This month we look back on
security vs autonomy (service) and how they illuminate data leakage and
potential privacy liability.
|
 Toby Considine
TC9 Inc
The New Daedalus
Contributing Editor
|
Recently,
Ken Sinclair suggested I reach back to some earlier columns that are
newly relevant for the Internet of Things (IoT). The buzz around
IoT is now white hot. This has attracted a host of fresh thinkers,
including many who think nothing has happened in Buildings before the
Internet of Things. One new entrant asked recently: “Yes building
automation is very interesting as an area. Do you feel that the
abstract nature of IoT is understandable to facilities managers?
If not, what do you think needs to be done in order to make it relevant
for them?”
I told him that
many facilities managers are among the most technically savvy in a
building. They are challenged by the conflicting requirements: energy
savings vs. comfort vs. status vs. tenant retention vs. security
(physical, network, IT) vs. branding.
It is frequent that
today’s more efficient building systems fight. Anyone who has an
instant hot water heater and a modern dishwasher can listen to them
fight as the low-water dishwasher (one more deciliter of hot water
because this load seems dirtier) starts and stops the instant hot water
heater.
Overly “intimate”
forms of integration make many IoT cloud integrations difficult and
expensive to maintain. Where efficiency means working on small
operating margins (for energy, or water, or…), the importance of fully
understanding how new technology interacts with the rest of the
building gets greater. Smaller operating margins on the grid make
creating a smooth, well-managed load profile more important. If you
plant to add a microgrid, then a well-managed load profile reduces the
capital-intensive generation and storage needed.
Cloud-based IoT
creates new and poorly understood privacy issues. I know, we have all
seen that people seem willing to give everything they know and do away
to Facebook—but even so, the fast-growing installation internet
ad-blocking technology is disrupting today’s revenue models. Today’s
privacy law seems to create policies whose effect is to make much of
our lives more frustrating. In the US, policies on medical privacy seem
to allow everybody except the patient [and family] to see the data; the
effect of privacy liability management is primarily to excuse bad
customer service.
I mention medical
privacy because it is only a matter of time until facilities managers
get legal responsibility for managing privacy in the IoT. What tenant
information can I share? What metadata about their comings and goings,
and about the equipment that they use and when must be obscured from
third parties. (More and more purveyors of digital power management are
able to recognize every piece of equipment in a building by its
electrical signature.) Will the cloud-based IoT provider warrant there
will be no privacy leakage?
Many IoT ventures
pitches today lean more on the potential value of the data-mining than
they do on value to the tenant/customer. I characterize this as “fun
for the boys, but hard on the frogs.” I don’t believe this path will
lead to long-term success. It is my guess that an entire generation of
cloud-based IoT that delivers less value to the customer than to the
data-miner will fail, and fail spectacularly.
Local autonomous IoT in buildings is something different.
(Cognitive functions, Autonomy, and Integration)
http://www.newdaedalus.com/articles/2008/3/29/cognitive-functions-autonomy-and-integration.html
Improved service
and improved performance can be more easily achieved at a lower cost
within smaller realms of the complete HVAC system for a floor. Better
customer service and tenant retention can be achieved by more modular
and tenant-responsive security systems.
(Bouncer or Prison Guard)
http://www.newdaedalus.com/articles/bouncer-or-prison-guard.html
Autonomous IoT can
become a way to enhance tenant service while reducing resource
requirements, by self-learning to anticipate and offer the right
service at the right time, but no other times.
This
model of the IoT as the local autonomous coordinator preserves privacy,
can increase efficiency, and enhances tenant service far better than
will the cloud. It does not intrinsically offer and means of
coordination. Coordination of services and resources will continue to
be more important as resource operating margins become slimmer, and
variability of resources becomes greater as they become based on
distributed energy resources (DER).
(The Building Services Interface: The foundation of smart building security and interactions)
http://www.automatedbuildings.com/news/jan11/columns/101231111303considine.html
The answer is
light, loose models of inter-system communication that balance resource
usage over time, based upon the *local* varying supply. Loose
integration reduces the complexity of specific integrations—we should
aim for autonomous IoT applications that do whatever they do, and
self-integrate into the larger building ecosystem with minimal human
involvement. Today’s models of “green buildings fail because they do
not scale—one can do twice as many buildings only if one has twice as
many engineers on the payroll. Autonomous IoT agents are our best
hope to get past this.
(Just-In-Time Power and Water at the Edge)
http://www.automatedbuildings.com/news/aug16/columns/160730103404considine.html
(Eight Agents for Energy)
http://www.automatedbuildings.com/news/dec16/columns/161116012909considine.html
Transactive
integration is the loosest effective form of integration we know.
Systems do not need to know how other systems operate (mechanisms),
they only need to know how they compete (or supply) common resources.
Set the limits on the resources available, either by hard supplies, or
by budgets, and let the autonomous systems work things out. The
spontaneous market order can replace non-scalable engineering
piece-work.
http://www.theenergymashuplab.org/
The opportunities
are immense. But they will not be built on merely hooking up every
sensor in the world to a cloud and somehow creating value by using
Hadoop.
Author's Note
“I wrote this column in mid-December. More recently, my son forwarded me this news on cloud-Iot and the law.
Police seek Amazon Echo data in murder case https://www.engadget.com/2016/12/27/amazon-echo-audio-data-murder-case/, a case that is sensational, but the third paragraph will be more important to IoT practiitioners…”
footer
[Click Banner To Learn More]
[Home Page] [The
Automator] [About] [Subscribe
] [Contact
Us]
