Will a new breed of IoT-based wildfire early warning systems coming onto the market make a difference in a world increasingly beset by flames?
Flames flick on the horizon and spread uncontrollably into the valley below. The blaze razes a cluster of outbuildings, then rips into a group of family homes, leaving in its wake smoldering houses, charred possessions, and ruined lives.
This has been an all-too-common scene in the Western United States during the summer of 2021. The air quality consequences from fires in the Western States are far-ranging, with smoke from the blazes drifting as far as Chicago and New York and causing major air pollution in those regions. Preventing fires is probably beyond the scope of technology, but mitigating the severity? That’s entirely possible. There is a small but growing number of IoT companies attempting to prove the point.
People debate the reasons for the growth in wildfires, citing everything from the obvious root cause of climate change to local origins such as 5G cell tower fires in Colorado or target shooting in Nevada.
Regardless of the causes, the effects of wildfires are devastating, and they are occurring far more frequently as we careen into the second quarter of the 21st century. The National Interagency Fire Center says that as of July 20th this year, 83 large fires have burned across the United States, and over 2,585,492 acres have gone up in smoke so far. Alarmingly, experts say that wildfires are responsible for around 20% of annual global CO2 emissions.
The Bootleg Fire is the wildland blaze du jour in the U.S., burning over 400,000 acres of forest in Oregon. Concurrently, the Dixie Fire in California has swept through more than 197,000 acres across three counties in the Golden State, and is only 22% contained as of Monday, July 26th, according to CalFire.
“On average, wildfires burn twice as much land area each year as they did 40 years ago,” says Climate Central, an independent organization of scientists and journalists researching and reporting on global warming.
So what can we do to prevent wildfire, asks Smokey The Bear?
The earlier a wildfire is discovered, the easier it is to stop or contain. When dangerous conditions are detected before a fire breaks out, authorities can prevent a disaster before it starts. Early detection of fires in forested regions, however, is extremely difficult. Before the advent of aircraft monitoring and satellite-based GPS technology, areas prone to wildfires depended on weather reports, an army of seasonal fire lookouts, and a great deal of luck to combat wildfires.
A few startups as well as other firms are developing IoT systems using wireless sensors to give early warning of wildfire activity. Companies delivering these systems currently include Dryad, LADsensors, and Seidor.
EE Times spoke to Carsten Brinkschulte, co-founder and CEO of Dryad Networks about how his company is introducing a solar-powered LoRaWAN-based sensor system for early detection. “A key part of our innovation…is that we added… a mesh network infrastructure to LoRaWAN. We basically added the ability for the base stations to talk to each other and do kind of Chinese whispers, one receives the message, it passes it on to the next and the next until it reaches a gateway that is Internet-connected… With this we can cover vast areas, like thousands of square kilometers, without each base station being connected to the Internet,” Brinkschulte told us.
The sensors can connect to border gateways that link to the Internet via LTE-M, or Ethernet connectivity to add a Starlink satellite dish. “Every border gateway is Swarm-enabled,” as well, Brinkschulte added.
The sensors are batteryless and have no user-serviceable parts. “They just hang off a tree,” Brinkschulte noted. Indeed, Dryad has even used the Internet of Trees as a tagline in some of its press materials.
The sensors integrate a Bosch BME688 gas sensor chip that detects the gas composition of the air — it looks at hydrogen, carbon dioxide, carbon monoxide. Then the Dryad sensor uses machine learning and edge processing to detect the combination of gases typical for a wildfire. “We get high accuracy and low failure rates,” Brinkschulte claims.
Catch a fire
Dryad is aiming for sub-one-hour wildfire detection. “When you’re looking at fire, the earlier you can catch it the better, right? As simple as that. And we can detect [a fire] before it is a fire, when it is still in the smoldering phase,” Brinkschulte says. This makes it much easier for fire services to deal with a problem before it becomes a wildfire disaster.
In a wildland-urban interface area that has roads as well as forests, you would need about 500 sensors to cover 10 square miles, Brinkschulte says. Once you get into wild forest expanses, the number of sensors needed would drop to about 250.
The standard price — before any volume discounts — is currently about $50 per sensor, the CEO says. This translates to around $25,000 for sensors to cover a wildland-urban interface area.
“Of course we’ll give volume discounts,” Brinkschulte notes. He expects the price of a sensor to drop to about $20 each, once Dryad ramps up to large-scale manufacturing, producing millions of units.
The startup also charges an annual subscription fee for the service, which constitutes about 10 to 15% of the hardware cost. So, this isn’t an IoT product designed for your average consumer (or even a farmer) on the wildland-urban edge. The CEO says that Dryad has been talking to governments, energy companies, and private forestry firms (which is to say, logging companies) as it introduces its product.
The CEO notes that Dryad can add other types of sensors to the network as it is rolled out. This, he says, will be of particular interest to the logging companies, as they can look at soil monitoring and tree growth as well.
It is still very early days for IoT-based early warning wildfire systems. They haven’t really been tested in the heat of the flame. These new IoT systems could potentially provide a new level of defense against wildfires, at a time when the world desperately needs it.
“It’s not going to solve the problem single-handedly, but we think it’s a very important addition to the puzzle that’s extremely complementary to satellite or camera-based solutions because we have this ultra-early detection capability,” Brinkschulte states.
This article was originally published on EE Times.
Dan Jones is a veteran reporter who has covered many segments of the communications market.