Wildfires are fast and furious.
With the unexpected hot weather, increasing wildfire risk in WUI, rising home construction costs, and the recent exodus of insurres from the California home insurance market, homeowners are facing significant financial risks if their properties are damaged by wildfires.
Homeowners need solutions to mitigate wildfire risks and damages. They focus on home hardening inside their properties and reducing their risk from distant wildfires that burn forest areas beyond their control. Early detection empowers homeowners to inform government agencies promptly, enabling firefighters to protect people more effectively. Homeowners must take timely actions as well since wildfires can spread as fast as 14 miles an hour! A faraway blaze can endanger you within a matter of hours.
We have received numerous inquiries from homeowners recently, including Eric. During our first Zoom call with Eric last Monday, he expressed his need to detect early-stage wildfire outbreaks within a 10-mile distance from his house. He had reached out to several wildfire detection solution providers, but their offers were costly, time-consuming, and did not meet his affordability and urgency requirements.
We listened to Eric’s concerns and inquired about the computing equipment and sensors he uses in his house. Eric mentioned having a router, broadband internet service, an iPhone, a MacBook, and several smart home devices such as thermostats and security cameras. When asked about his security cameras, Eric specified using two Reolink RLC-820a fixed cameras – one outside the garage and the other monitoring the front door garden, with a clear view of the large forest areas 10 miles away.
We assured Eric that we would explore potential solutions for him and scheduled a follow-up meeting for two days later.
We’re exploring making use of Eric’s existing equipment to detect wildfires and avoid new hardware and installation time. In our research of the webcam market, we discovered that most home security cameras are walled-garden systems and do not support third-party software. However, Reolink stood out as an exception, with certain camera models, including the RLC-820a, offering an API for accessing real-time images. This finding led us to believe that a simple software solution running on Eric’s MacBook could utilize his front door RLC-820a camera to provide image input to our LookOut wildfire detection SaaS at regular intervals.
Taking action, we promptly contacted Reolink, ordered an RLC-820a camera, and developed a prototype for our solution. Upon receiving the camera, we built our MVP (minimum viable product) prototype.
During our second meeting with Eric on Wednesday, we explained the functionality of our solution and conducted a real-time demonstration using our RLC-820a camera. Our solution only requires messenger.sh, a simple shell script to be run on Eric’s computer.
Below is the source code of messenger.sh:
#! /bin/sh
# A shell script to get live jpeg image from the Camera and send it to LookOut Wildfire Detection SaaS. Developed by @roboticscats
# Camera: Reolink RLC-820a
# Last update: 20230716
# Your LookOut Wildfire Detection SaaS camera information
lookoutCamName="Reolink-820a"
lookoutCamEndpoint="https://…"
# x is the variable to keep track of the while loop
x=0
# cycle is the number of cycle (get image from camera, post image to lookout). Default values is 5 for testing purpose, enter 86400 for Premium Plan and 43200 for Standard Plan respectively.
cycle=${1:-5}
# 30 seconds for Premium Plan, 60 seconds for Standard Plan
interval=60
# start and startsecond are the time this script starts
start=`date -Iseconds`
startsecond=$(date +%s)
while [ $x -lt $cycle ]
do
# get still image from Camera and save it temporarily to the file temp.jpg. Use your actual device IP address and user credentials.
curl "http://192.168.1.100/cgi-bin/api.cgi?cmd=Snap&channel=0&rs=123456&user=wildfire&password=detection" -H "content-type: image/jpeg" -s --output temp.jpg
# send the image temp.jpg to your LookOut camera endpoint
curl "$lookoutCamEndpoint" -H "content-type: image/jpeg" --data-binary @temp.jpg -s --output result.txt
x=$(($x+1))
sleep $interval
done
# the following codes provide a progress summary
end=`date -Iseconds`
endsecond=$(date +%s)
echo "$x images are send to LookOut camera $lookoutCamName in the last $(($endsecond-$startsecond)) seconds between $start and $end." > summary.txt
cat summary.txt
We assisted Eric in setting up a LookOut account and customized the shell script to work with his camera. As a result, his front door Reolink RLC-820a camera was able to send live images to LookOut, detecting wildfires captured in the scenes. Eric has been using our LookOut service since our second meeting. He expressed gratitude for being able to utilize AI wildfire detection to protect his family, while we were delighted that our MVP facilitated a quick understanding of customer needs and progressed our discussion toward customer adoption.
“Thank you RoboticsCats for your prompt and friendly service,” said Eric.
For more information on how to use the Reolink security camera with LookOut wildfire detection service, please contact us at info@roboticscats.com.
