HP Labs gets smart about energy use – starting with its own main building
Contributed by Simon Firth, freelance technology journalist
Next to every pair of overhead lights in HP Labs’ headquarters office – in a facility known as Building 3 – you’ll find a small white disk, the size of a smoke alarm.
Each disk connects to a dimmer that controls the room’s lighting nearby and inside each disk is a tiny wireless transmitter that connects to a central server. Thanks to these commercially available third-party devices, it’s possible to remotely adjust the lighting at a building-wide scale down to the illumination of a single desk.
Beyond that, however, every disk contains a light, temperature and motion detector. This allows for a dramatically more energy-efficient and automated configuration, notes Geoff Lyon, the HP Labs researcher who led the new installation alongside HP’s Global Real Estate (GRE) team.
Monitoring Dashboard under development by HP Labs, Puerto Rico Solution Factory, and Energy Solutions teams
“The old lighting system had zones of lights that would be controlled through a single switch,” he explains. “So large areas would be lit even when light was only needed for one person in their cubicle. Now the energy we use for lighting is much more in line with the activity within the building. If you arrive early in the morning before anyone else, for example, each individual pair of lights will brighten as you walk towards them and then dim behind you.”
The new system has already significantly reduced HP Labs’ energy consumption in Building 3, Lyon says. But the ambitions of the HP Labs Sustainable Ecosystems Research Group to which Lyon belongs go much further. They’re now adding hundreds of additional meters and measurement systems to track all the utilities used in Building 3, including gas, water and even waste.
In all, the Labs team expects to install close to a thousand smart meters, each of which will feed real-time information about when and where energy resources are consumed in the building.
Labs’ researchers suspect they’ll find plenty of places where the energy use is greater than the need, says Amip Shah, principal investigator for the Resource Management as a Service (RMaaS) project, of which the Building 3 installation is a part. With the monitoring infrastructure in place, says Shah, “we can begin to ask, what are the dependent parameters that influence the consumption of energy in the building?”
“And that’s a non-trivial problem,” he adds, “because we already have tens of thousands of parameters in our data system, sending information every few seconds. But with this holistic view, we’ll be able to fine tune the lighting, heating, and cooling of the building to much better match the actual use of the space as it varies through the day and through the year.”
It’s likely that a commercial installation of such a service wouldn’t need as many sensors as Building 3, Shah notes. But part of the point of the project is to establish exactly how smart a “smart metering infrastructure” needs to be before it can have a real impact. Beyond that, the Sustainable Ecosystems team is looking to create a variety of sophisticated tools that can both model energy consumption and suggest corrective actions in real-time based on that information.
Monitoring Dashboard under development by HP Labs, Puerto Rico Solution Factory, and Energy Solutions teams
While the Building 3 monitoring network was built primarily as a research test bed, the savings that it’s already created have convinced HP’s Global Real Estate team to evaluate similar projects for other buildings on HP’s Palo Alto campus and to explore opportunities at HP sites further afield. In fact, the GRE energy team involved in the Building 3 lighting project and the HP Labs research team are jointly proposing new shared initiatives in a growing collaborative effort. And, teams across HP are collaborating to enhance the deployment of energy and sustainability monitoring software across the company’s facilities.
Significantly though, once you go beyond the building level, it isn’t just the building’s owners that stand to gain from the information reaped from an RMaaS installation.
RMaaS data can also be used by utility companies to predict resource demand, which helps them avoid the problem of under- or over-provisioning of the electricity, gas and, water they supply. Utilities might even help large organizations install RMaaS systems on their campuses for just that reason. The smart meter systems could also help utilities determine which incentives would have the most impact on particular customers. “That’s actually a really hard problem to solve today,” notes Shah.
Fundamentally, he says, “too often people just don’t know how many joules of available energy they’re using. We want to help organizations figure out a pathway to a meaningful and achievable energy reduction target, and at the same time offer ways to measure whether they’ve been able to meet it – from the building to the campus, and even up to the scale of entire cities.”
