Never Miss a Beat: Get a snapshot of the issues affecting the IT industry straight to your inbox.
We Need Multifaceted Tech Now to Curb Extreme Wildfires
A daunting new era of wildfires and an aging energy grid require that we combine technologies, including hardware, data tools and AI models, to extinguish the threat.
4 Min Read
LEBLOND Catherine via Alamy Stock
Extreme fire weather fed by heatwaves of greater intensity and droughts of longer duration is scaling fast and furious globally. Amid climate change, wildfires are expected to increase by almost 30% within two decades and, by the end of the century, up to 50%. Complicating matters, we’ve learned that last year was officially Earth’s hottest year on record. In North America, “megafires” that burn more than 100,000 acres may be worse than at any other time in recorded history, and the more acres burned correlates with higher temperatures. Not least, more smoke that’s traveling further is spewing all-time high wildfire air pollution.
No single solution can extinguish this threat. Only a combination of efforts and multifaceted technologies, including AI-driven tools, can get humanity in front of the inferno.
Ignition From the Grid
The energy grid offers a compelling example of wildfire risk -- and how multifaceted technology can address prevention, detection, battle, and recovery. While wildfire causes include arson, carelessness, and natural catalysts like lightning, data from the National Interagency Fire Center shows that federal, state, and local fire services confronted 32,652 powerline-ignited wildfires from 1992 to 2020. About 80% of the US grid is still above ground -- exposed and old.
In fact, 70% of transmission and distribution (T&D) lines are well into the second half of their lifespans. At the same time, we’re putting more load on our electrical transmission systems and the risk of sparking is ever-present. When components ignite, it’s often not detected until a fire is blazing. The good news is that we can address the danger.
Remotely Monitored Spark Prevention
On grid lines and equipment, surge arresters protectively divert excess voltage from electrical surges or lightning to the ground. But they’re far from infallible. When fortified with robust hardware, the risk of fire decreases. A spark prevention unit (SPU) monitors the surge arrester’s current and thermal load; if there’s an overload, it interrupts the current flow and disconnects the surge arrester. This prevents any arcing, sparking, or ejection of hot particles that can start a wildfire.
These SPUs are specifically made for areas at high risk of wildfires and include a wireless indicator with long-range transmission (LoRa) capability that monitors the SPU remotely and takes action fast. The indicator transmits data about the time of any tripping event and geographical location, as well as SPU status. Because they are safe and effective, they are approved for vegetation clearance exemption in California, with hundreds of thousands installed in irregularly inspected and difficult-to-access wildfire-prone areas. But more technology of this kind is needed across the country.
Image-Based Inspections Backed by Sophisticated AI/ML
Drones need wider reach too. When deployed in repeated patterns to capture images and data, their analytics-based inspection software can quickly identify defects in powerline and grid assets by leveraging customized AI/ML models. The software can automate defect assessments and instantly analyze thousands of multi-angle images from different sources -- including photographic, video, LiDAR, thermal, and satellite images. It conducts identification, cataloging and health evaluation, and its computer vision algorithms and ML determine the failure potential of granular assets like dampers, ceramic disks, pins, polymer insulators, wooden poles, and more. As needed, humans can enter the loop with inputs that further train AI models.
An effective data system must be in place to process and analyze the massive amounts of visual data collected, along with sensor data and historical data, to get the most accurate view of what’s happening, and what likely will happen, on the ground. In detecting and fighting wildfires, fast data processing that gets the right alerts to the right people who can make proactive decisions is vital.
Rigorous Predictive Models for Air Quality
Beyond stopping the flames, understanding the impact of the smoke on people and animals -- and the levels and distribution of smoke -- is critical too. The red pall from Canada’s extreme wildfires that covered the northeastern US last year and surprised populations will become increasingly common. In fact, a second round of these wildfires and the air quality issues they create are likely this year.
Communities need end-to-end systems for ingesting air quality data from IoT sensors, meteorological sources, and other tools, then automatically preparing, validating, and blending that data. Fine-tuned AI/ML and advanced data analytics can be applied to build rigorous predictive models for air quality.
Other multifaceted technologies are on the horizon: new mapping software for vegetation growth, encroachment, and prediction; satellite technology that can pinpoint the unusual heat of wildfires as they break out; and solar-powered sensors on trees measuring gas and humidity while gathering data from infrared cameras. AI can even help battle the fire itself by predicting its movement, so firefighters can determine the best placement and size of firebreaks.
The cost of extreme wildfires is soaring. According to a CBO report, “between 1989 and 2020, the five-year moving average for federal spending on wildfire suppression more than tripled in inflation-adjusted terms.” But the cost to people and the planet may, in the end, be what finally spurs collaborative action using all available technologies and data.
About the Author
You May Also Like
More Insights
