Related
Recommended
Engagement
  • Author Author: cstanton
  • Date Created: Date Created
  • Last Updated Last Updated:
  • Views 3276 views
  • Likes 10 likes
  • Comments 13 comments

Color Sensor-Based Water Quality Tracker: DIY Environmental Monitoring -- Episode 621

Katie delves into the potential of color sensors to detect changes in the color of water, indicating variations in water quality. Join her as she navigates the technical challenges and solutions involved in creating a low-cost and effective water quality monitoring system that can benefit environmentalists and researchers alike!

Watch the Video:

In this video, Katie explores the concept of sensing water quality using a color sensor. Her project is part of the element14 Community's Summer of Green Tech, which features discussions and projects about innovative ways to use electronics for environmental purposes, including water quality monitoring. Katie starts by sharing her initial idea of using a color sensor to assess water quality, particularly focusing on the turbidity and color of water.

colours

She starts by connecting a Grove I2C color sensor to an Arduino board in order to test its ability to detect the water’s color. Katie experiments with cups of red-colored, green-colored, and clear water to see how the sensor reacts. This preliminary test demonstrates that the sensor can differentiate between different colors, laying the foundation for potential water quality assessment.

with electronics

Katie details her plan to address the challenge of measuring water quality in a stream without disturbing the flow of the water. She plans to use a pipe to create a controlled flow environment for the water. She also raises concerns about protecting the sensor from water damage and avoiding potential light interference with the sensor.

Dipping

Katie explores the process of encasing the sensor in clear resin to make it waterproof, while still maintaining its ability to accurately detect color. She describes the steps involved in creating a silicone mold and filling it with optically clear resin to protect the sensor.

Sensing

The video concludes with a successful test, where the sensor detects the differences in color between clean water and water mixed with coffee grounds. Katie emphasizes that this project has laid the groundwork for a potentially low-cost water quality sensor that could be widely used, and invites community members to contribute to its development, particularly on the coding side.

Success

Overall, this video showcases the practical steps and technical challenges involved in using a color sensor for water quality assessment, with the potential for broader applications in environmental monitoring.

Supplemental Content

Bill of Materials

Product Name Manufacturer Quantity Buy Kit
Arduino MKR WiFi 1010 ABX00023 ARDUINO 1 Buy Now
Color Sensor Module SEEED STUDIO 1 Buy Now
White LED Strip LEDXON MODULAR 1 Buy Now
Clear epoxy potting and encapsulating compound MG Chemicals 1 Buy Now
Silicone resin Wiska 1 Buy Now
 

Actions

  • Interesting project  . The project idea is really innovative and can be helpful for detecting water quality,

    I am thinnking of my past readtest of Thingy:53 on element14 where I had a color sensor libray but it also have air quality sensor. So these two sensors can perhaps be combined to detect water quality. The air quaality sensor can measure volatile gas compounds and co2 concentration arround it. tVOC and Co2..etc. Combining ML algorithms would also make sense.?

    What do you think?

  • in reply to beacon_dave

    Oh yea, didn't think about organisms thriving on the light. A UV light is a good call though, wonder if you can get LED strips of UV light likes you can in usual LEDs, I'd guess so, but might be slightly more challenging to find them than white and/or neopixels are. Then it could be run down the other side of the tube.

  • in reply to hifromkatie

    I suspect that some organisms will thrive on the LED illumination. It may be possible to sterilise with a UV light source though to prevent build up around the sensor.

    There are various antifouling treatments for marine applications but not sure if anything is optically clear. Perhaps you might find something aimed at the aquarium market.

  • in reply to beacon_dave

    I like the coffee monitoring idea, it's always disappointing to get a sip of weak watery coffee, I could check before taking a sip!
    I did wonder about build up, but couldn't think of a self cleaning method, other than making sure there wasn't crevices or places for things to build up. It why I did the led as a continuous strip down the length, but the sensor is now a weak point. I wonder if encapsulating it in the tube rather than sticking it through a hole in the side would be better, but ultimately I think it'll need some manual wiping out periodically.

    Thank for the link, I'll look into that, as I'd be useful with all colour sensor projects

  • in reply to DAB

    Thanks, it would be just a first line test to show you might need a more in depth check. Interestingly our water company have message everyone today to say we need to boil our tap water before drinking until they inform us otherwise!

  • Nice project Katie.

    I like your solution, you can do a reasonable job of measuring clarity, which is always a good first test.

    If you start with clear water, then you can usually just boil it to get rid of the biological nasties.

    It probably won't help you with some chemical contaminants, but you will need to go to spectroscopy for those. 

  • Another interesting project idea and looks great for those STEM type activities.

    As a bonus you can now measure the colour quality of your coffee Slight smile

    One thing perhaps to note is that if you leave this out in the wild for a prolonged period it is likely to get pretty dirty as nature decides to move in, which will then affect the readings. So the design should perhaps allow for a standard bottlebrush to easily clean the sensor and light parts. I suspect that the amount of gunk will vary depending on the flow rate of water through the pipe (self-cleaning?) and how much sunlight can reach it (algae). One would hope that the inside surface of potable water pipe would minimise this but I suspect that some resins may attract more nature than others.

    May also want to place a coarse filter over the ends to prevent aquatic critters from taking up residence, although that also is likely to get gunked up over prolonged periods.

    I'm not sure if the sensor library incorporates it or not but there is usually a matrix transform function that needs to be applied to the sensor data to produce more meaningful calibrated results, especially if you want tristimulus values / chromaticity coordinates to be able to use the captured data alongside other data. Luckily there is a 2005 whitepaper by TAOS that takes you through the process.

    https://ams.com/documents/20143/36005/ColorSensors_AN000518_1-00.pdf/16f9c551-6778-1a39-d775-e53ed5b9af3f

    I recall from a colorimeter project done with one of the early versions of those sensors, that you can get surprisingly good colour accuracy once calibrated.

    Another feature that might be worth considering in the design is that you may find that you need to adjust the sensitivity of the set-up depending upon the normal clarity of the water being sampled. Being able to vary the intensity of the light source or adjust the distance between the sensor and the illumination source such that more or less light can reach the sensor to keep it within the optimal sensing range. If you compare the clarity of dark peaty water with tap water, then that's a huge range. 

    It may also be worth considering having a light sensor attached directly to the illumination source as an illumination reference.

  • in reply to kmikemoo

    It took a fair amount of playing around to get it right.

    The oven was also super useful, as waiting ~2 days for it to set at room temperature meant there was too many chances for it to get nudged and then set wrong.

  • in reply to dougw

    I hadn't thought of that, but it would save having to water proof the LED strip, it was much more translucent than I thought, but by the time I realised the LEDs were always in there.

  • in reply to baldengineer

    Yes, that was the aim of the project, tell you if something has changed and if it requires more investigation. Think it'd pick up algae well.

    Wish I could remember the guys name who gave me the idea for this project, but his aim was a sensor that would be cheap enough to deploy everywhere, that could tell them if there might be a problem that they need to bring the expensive water monitoring equipment to that location to investigate.