CBC reports that with 4,370 kilometres of pipes carrying clean water across the island, Montreal's water main network is one of the largest in Canada. At least a third of those pipes have reached the end of their lifespan, and another third will by 2020. In 2011, there were 29 breaks per kilometre of pipe. It was down to 20 breaks by May of last year. According to Ronald Gehr, a civil engineering professor at McGill, the main factors include the age of pipes, pipe material, soil conditions, and freezing water within the pipe.

Pipe patroller is an IoT system that looks at deploying sensors along stretches of water pipes, providing a wealth of data about humidity, temperature, and gas pressure about the pipe and its surrounding area. This up-to-date time-series information allows for more accurate predictions of a pipe’s lifespan, changing along with the environment. We can use this data to evaluate the condition of pipes new and old, and opens the door for a proactive approach to repair rather than being simply reactive.

https://www.cbc.ca/news/canada/montreal/montreal-water-infrastructure-problems-1.4660551

Team Grafana

Matthew Gaiser, Christopher Lai

Inspiration

Smart cities are all about collecting data and understanding the state of the city. Montreal has been investing effort and money into addressing the many water leaks across the city, and a smart network of devices will allow efforts to be more proactive, and not simply reactive.

What it does

We read sensor data from an arduino sensor package, processing that information and rendering it with Grafana. We then write queries and displays that take advantage of the time-series nature of the data to slice data across time, and identify the nodes most at risk.

Tech Stack

PipePatroller is built with Arduino, Grafana, Google Cloud Platform, Python, MySQL, and love. Arduino modules used include:

  • An ESP8266 wifi module
  • Two temperature and humidity modules
  • An air pressure module
  • A 9 volt battery for mobile power
  • A capacitive soil moisture sensor
  • And an Arduino Uno to serve as the base

Challenges

  • The WiFi network blocked ports required for the sensor to upload data to the web
  • Memory limitations on the Arduino when sending requests mangled requests with no hints
  • Formatting and query issues when creating visualization panels
  • Inconsistent GPS coordinates between Google Maps and Open Maps (visualization)
  • More aware and appreciative of the challenges associated with arduinos and communicating across networks
  • Better idea of domain knowledge related to pipe corrosion and replacement.

Accomplishments that we are proud of

  • Successfully integrating our hardware and software components
  • Getting the google cloud instance running smoothly having been the first time working with the platform.
  • Strong organization of the API and required data before coding even begun

What we learned

  • More aware and appreciative of the challenges associated with arduinos and communicating across networks
  • Better idea of domain knowledge related to pipe corrosion and replacement.

What's next for PipePatroller

  • This technology could be adapted for general ground monitoring, specifically areas which are close to foundations or in areas which are experiencing significant changes due to climate change.
  • Power management technology. Theoretical estimates have the device lasting up to a year with a 2500 mAh battery if the required changes are made for power conservation.
Share this project:
×

Updates