-
-
Smart Flood Sensor Using Arduino
Inspiration:
When my family and I got stranded in city streets due to heavy downpour and storm, I could see traffic moving at a very slow pace, on-ramps getting closed due to floods, traffic getting rerouted, and cars loosing traction in small flood puddles on freeways. Basically, it was a bad evening to drive outside but since it was a school and work day, we could not avoid driving and had to reach our destinations on time. It was this that inspired me about doing research on what the modern technology can do to solve flooding problem. I became passionate on solving it when I found the reasons and the strategies that could help mitigate the problem.
What it does:
Smart Flood Sensor is a device that can be used in a network of flow sensors in a city's drainage system to collect, analyze, and predict rain and storm water flow data. Specifically, the device is able to calculate water flow with 90% accuracy, upload the data periodically to the IoT cloud, and send text alerts to nearby utilities crew if flood levels reach a certain threshold. The analysis software is able to predict flood at a drain, given the historic and real time flood level buildup at adjacent drains. It also recommends optimal drain pipe sizes needed to fix the choking points. The software component is serviceable wirelessly within 25 feet radius.
How I built it:
My device uses a Hall Effect based water flow meter, Arduino 101 microcontroller, GSM/GPRS shield, a Bluetooth link for UART connection, and a rechargeable battery (10K mAh), all enclosed in a UL rated waterproof enclosure. It has low power requirements (~20mA idle, 40-60mA operating), small form factor (4.5"x4.5"x2"), and is non-obstructive to water flow. My software running in this device calculates flow rate, connects to thingspeak.com servers, and uploads data in 15 sec intervals. Analysis phase uses my MatLab software and my graph theory algorithm to predict flood levels in the drain network.
Challenges I ran into:
The first challenge I ran into is to find a way to mix water with electricity in the same device. I had to make the device completely water proof and test it with immersing the device completely under water for over 24 hours at a depth of few feet to test the leak as well as water pressure. Second challenge was to get a enclosure that is not only water proof but also have a form factor that is suited for drain openings.
Accomplishments that I'm proud of:
I tested my project by installing devices at five different drain openings in my neighborhood streets. During the recent rainy days, these devices successfully recorded water flow data and uploaded them to the IoT servers in 15 sec intervals. In parallel, the analysis software analyzed data every 15 minutes. One of the devices sent a text alert that it has detected an excessive water flow (>350ml/sec) at a drain. HOA was notified to increase pipe capacity at that opening. It also identified a drain where flow was gradually decreasing while other drains showed higher flow rate. HOA was alerted in real time and it was found that debris was the cause. This device saved a flood occurrence in that street. In another case, HOA was alerted of a leaking sprinkler.
What I learned:
Regular flooding in city streets proves that the city planners are not connected to what is happening under the ground. My Smart Flood sensor, a low cost device, can not only help identify drain openings in city streets that are getting clogged in real time, but can also determine areas where more drain pipes are needed. Arduino 101 is very well suited for this IoT project because of its form factor, easy access through bluetooth capability and the cost. With the availability of cloud servers, the use of IoT is even more feasible.
What's next for Smart Flood Sensor:
I want to reduce form factor of my Smart Flood Sensor device. For this, I will be trying to replace the larger components (such as GSM shield) with a smaller low power ones. Once the form factor is reduced, I will attach other sensors to the device such as - accelerometer (to prevent theft or detect physical dislocation of the device), temperature sensor (to proactively service the device if temperature is out of range) and humidity sensor. I already showed my device to Cupertino City Public Works department and they gave me a good feedback which I will be incorporating into the device.
Log in or sign up for Devpost to join the conversation.