Working on the cutting edge of service-oriented architecture (SOA) and event processing software, IBM researchers have helped develop a virtual market where home clothes dryers can bid on the price of the electricity they use.
This is not a futuristic plan or proof of concept, sponsored by the U.S. Dept of Energy it is operating in homes in the Pacific Northwest, which announced the successful completion of pilot projects Wednesday at a teleconference.
"These demonstrations clearly show that there are no significant technical hurdles standing in the way of wide scale adoption of demand/response automation using pricing signals and grid-friendly appliance controls," said Ron Ambrosio, senior technical staff member at the IBM T. J. Watson Research Center, where the SOA technology for the project was developed.
As part of the Olympic Peninsula project, 112 homeowners received new electric meters, as well as thermostats, water heaters and dryers connected via Invensys plc home gateway devices to BPM and event processing software running on IBM WebSphere servers.
"IBM was the overall system architect and integrator of the real-time market price and control system that brought together all the smart devices, the real-time event software, and the advanced analytics to provide the homeowners with detailed information about their energy usage and associated costs," Ambrosio explained.
The Department of Energy's Pacific Northwest National Laboratory worked with the IBM Watson researchers to develop the real-time market price and control system that makes home electricity customers active participants in managing the power grid and their monthly utility bills, he said.
"Using software from IBM and intelligent devices installed in customers' homes and at certain power suppliers the project created a virtual marketplace," Ambrosio said.
As he explained, customers would be able to set limits on how much they would pay for electricity to run a clothes dryer, for example, and electricity providers willing to transmit power at that price would be alerted over the grid and could sell the electricity to the dryer. Also, during peak power usage, thermostats would be raised or lowered in the home to reduce electricity consumption and avoid blackouts. This would all be done based on parameters the home customer set, such as keep the house at 76 degrees during summer days, but if there is a power shortage raise the thermostat up to 80.
On one side, consumer devices were bidding for power based on how much they were willing to pay based on customer preferences that had been set ahead of time, the IBM researcher explained. On the other side, suppliers entered bids automatically from their electricity generators, for example, based on how much it would cost to start up and run them.
On average, consumers who participated in the project saved approximately 10 percent on their electricity bills, according to the Department of Energy report on the project.
"IBM helped create this event-driven service-oriented architecture to support sharing of information that was critical to the real-time pricing control system," Ambrosio said. "While event-based processing and SOA are not household terms they are really the under the hood capabilities that provided the backbone to support this real-time market pricing system."
The project made cutting edge use of not only SOA, but business process management (BPM) and what IBM researchers call event-driven programming.
"One of things we wanted to develop and demonstrate in this project was making things interoperate all the way from the device on the wall to a business process sitting on a server that was implementing a real-time market," Ambrosio said. "That's a pretty groundbreaking thing that happened there. We linked a business process in real-time directly to physical devices."
Asked how important SOA technologies were to the creation of the virtual marketplace, he said the service-oriented approach was at the heart of it.
"This goes beyond what we even have today with service-oriented architecture, which allows us to glue everything together at the enterprise level," Ambrosio said in response to a reporter's question. "What we've been using in this project and other intelligent utility network activities is new work coming out of IBM research, which we refer to as the event-based programming model that extends our SOA environment out into the sensor and control system environments and deal with time sensitive systems like this. That was basically what made this happen."
Ambrosio noted the energy bill recently passed by Congress and signed into law by President George W. Bush, specifically calls for the development of a standards-based interoperability framework for energy providers, appliance manufacturers and consumers to use.
"The interoperability framework in the new energy policy act, it's what SOA is," Ambrosio said. "It's what event systems are."