Newswise – Multi-channel sensing can help in emergency situations.
Smokey Bear has a lot of great advice on preventing wildfires. But how do you stop one that has started before it gets out of hand? The answer may lie in pairing multichannel sensing with advanced computing technologies offered by a new platform called Sage.
Sage offers a unique blend. This combination includes multiple types of sensors with computing “on the edge”, as well as including machine learning Algorithms that enable scientists to process vast amounts of data generated in the field without having to transfer it all back to the lab. “At the edge” computing means that data is processed as it is collected, in the field, while machine learning Algorithms are computer programs that train themselves how to recognize patterns.
Sage is funded by the National Science Foundation and developed by the Northwestern-Argonne Institute of Science and Engineering (NAISE), a partnership between Northwestern University and the US Department of Energy’s Argonne National Laboratory.
Researchers using Sage recently completed a demo in which they successfully monitored a controlled burn — in which a plot of land is carefully burned as part of environmental management — of a portion of a tall Konza meadow in Kansas. Sage’s deployed advanced electronic infrastructure, which enables real-time detection, monitoring and analysis of burned area, can provide scientists and natural resource officials the ability to respond to wildfires with multi-tool data that is rapidly analyzed.
“When it comes to bushfires, time is absolutely of the essence,” said Rajesh (Raj) Sankaran, Argonne computing scientist and NAISE Fellow. “Often, there is no time to transmit data from the field – where high-speed connectivity can be an issue – to the lab. With Sage, we get the relevant information we need when we need it.”
Controlled burning in the Konza meadow area gave the researchers a large data set – nearly 60 DVDs – filled with information about the evolution of smoke and fire. This data can be used to educate a machine learning An algorithm that can make further determinations of the behavior of other fires in real time.
Following the success of the Sage Network in Kansas, there are future plans for the network to be deployed in California, Colorado, Illinois and Texas as part of a network led by the National Environmental Observatory Network (NEON). Ultimately, the researchers hope to create a continent-spanning network of smart sensors that could use Sage’s technology. “NEON is developing a mobile deployment platform that can complement onshore and aquatic sites across the country,” Sankaran said. “Sage could play a supporting role in many different settings across the United States.”
Technologically speaking, Sage is based on an open source wireless sensor platform called Waggle, developed and funded by Argonne. Waggle leverages emerging technology in processors, sensors, and low-power cloud computing to build robust, reliable sensor nodes that can effectively analyze and respond to data. “Essentially, Waggle is the foundation that Sage is using,” said computer scientist Argonne and NAISE co-director Pete Beckman, who helped with Waggle and Sage. “It’s as if Waggle is a cell phone, and Sage is the network the phone uses to communicate as well as the apps running on it.”
According to Beckman, the team is also seeking another research partnership with a researcher at the University of Oregon, who is working with the Federal Emergency Management Agency to build a series of monitoring stations in the Pacific Northwest. Beckmann hopes that by including Sage, these monitoring stations can add functionality in anticipating wildfires and other natural disasters by monitoring the environment.
The development of Sage through NAISE brings the scientific strength of two major research institutions. “The partnership between Northwestern and Argonne has long been fruitful of pivotal discoveries, and Sage is only the latest achievement that can make a difference for so many communities,” Beckman said.