Flood Network – Level Sensor Network

Hi all,

This is my first post on the board, but had to join in the conversation as I love what you guys are doing.

A couple of years ago we started experimenting with low-cost sensors and ultrasonics to measure water level. This year we launched a startup with some seed funding to get a network of water level sensors built. Various things have happened including trying to find a suitable business model to sustain it whilst still allowing the citizen science/crowdsourcing aspect to develop.

Our website (http://flood.network) is fairly light on technical details as it’s more about the end result (public flood map/private data products) but there is some history at http://oxfloodnet.co.uk which links to all the github code and a background blog post.

The hardware development has slowed over the past 9 months and I’m feeling the need to fix some problems. Some of the issues we get:

• Temperature sensors DS18b20 have mostly failed
• Distance readings are fluctuating slightly on single readings
• We’re using MB7060 which may not be as accurate as MB7389
• Maxbotix transducers comprise 80% of the cost of the devices and don’t have a great price at low volume
• Wireless range is terrible with our current radio and we’re planning to move to LoRaWAN as soon as possible

On the positive side, we have an IP67 rated device a few inches square, a reed switch to allow a reset without opening it, about 1 year battery life off a 4.7v lithium cell and the backend to process and log the data in large quantities.

Our aim is to bring all possible water level and flow sensors into the flood network to allow people to share their information quickly and easily. These can be user-installed, part of our commercial product, a third-party sensor or live data from an API (such as the one provided by the UK’s Environment Agency).

I can share more info and things we’ve learned from deploying sensors in the wild, but I thought I’d introduce myself. Hello!

Ben