Physical Computing Project Proposal – Well Tap

In ITP, Physical Computing

After some consideration, I’ve decided to move away from the shower sensors. Instead, I’m looking to focus on a network of sensors that will quantify peoples daily and weekly domestic water usage. The project is called Well Tap (WIP).


Executive Summary

Water waste is a serious problem, but unfortunately, even people that are aware of this continue to use way more water than they should. Because there are not tangible examples of this problem in day-to-day life (unless you live in California and Texas for Americans)  people often don’t even think about their behavior with regards to water. By providing people with clear data on their daily and weekly usage, and comparing that data to average numbers for the various household sizes, users will immediately get some context on their water usage. While trying to beat relative benchmarks will be a driver, I think people will be encouraged to beat themselves most importantly. The goal is not to make anyone feel bad or point fingers, but simply to make people more conscious as I believe that people are inherently good. When presented with reminders of the positive impact they could potentially make, I think this device could influence people to make small efforts that make a big difference.

For future iterations, there would be opportunities to also compare someone’s data to that of your neighbors and specific communities. It’s very important that data is compared to similar profiles as there’s not much sense in comparing one person’s water use in New York to a family household’s water usage in Arizona. At least as a first step, I can compare a user’s data to the average consumption of a corresponding household size.

Product info

A user would fasten/screw-in this device on their various water outputs at home (i.e. kitchen faucet, bathroom faucet, shower). It would screw in to the end of a faucet, and for the shower, it would screw in between the water output and the shower head. I need to use an adapter to make the 1/2″ diameter of the flow meter compatible with the standard faucet 1 1/4″ diameter.

The goal is to make a cylindrical encasing that has an LED indicator when the device turns on. I’d like to make it so the device only turns on if the flow meter is outputting readings. (Question: is there a part that controls turning on and off a circuit based on a sensor state or is this done programmatically). The blue light turning on would be clear feedback that the device is reading your water use. It serves as a friendly reminder to be mindful with your consumption. If I’m able to, I could add a secondary LED that could start flashing red after you pass a certain threshold of “a lot” water being consumed in one use.

Initially, I wanted to make to make the sensors battery powered and include all the components in a contained casing that would fasten to the water output. However, after a resident meeting, that approach may be too ambitious and actually not ideal as the wifi functionality would pump the battery life in a matter of hours. This is something that you would want leave “on” so it could continually track your water usage. Therefore, for this first run, the sensor casing that attaches to the water output will  simply have the sensor and LEDs, and it will connect to the arduino via wires that are neatly run along the faucet spout. The arduino would be hidden somewhere under the faucet and plugged into a wall outlet. This clearly introduces a couple of challenges when it comes to powering these sensors in different settings (i.e. there are no outlets by showers necessarily), but I want to first get this up and running in one environment and then I can think about scaling.

I’d like to create three of these devices that all communicate with the software interface. A household may potentially need more than three but it’s a good place to start. Also, there are things within the household that I won’t be tracking on the first version like the dishwasher, washing machine and toilet. All of these things consume a lot of water and  I’d like to get to a point where I could track them. However, in an effort to make this project feasible within the timeframe I have, I will focus solely on faucets and showers for now. In the following section, I will suggest a couple work arounds to make sure that the daily water usage is not too skewed by these omissions.

Software

There are two different directions I could go with the interface. I could either build it out in Processing or build it out using HTML, CSS and Javascript. The ideal would be building out a website, but I may not have enough time given that I have yet to learn Javascript yet. I will investigate more in the coming days.

When a user first loads up the interface, they will be prompted to say how many people are in their household. This will allow me to compare the user’s data to benchmarks of similar household sizes.

There’s one very important omission from my project. My system of networked sensors is not tracking some of the household appliances that consume the most amount of water—the toilet, dish washer and washing machine. In order to make the data I’m collecting from a user’s sinks and shower relevant to the benchmarks we’re using, I need to include the amount of water they’re consuming using this appliances. It won’t be feasible to add these appliances to my network of sensors given time constraints but I can have them enter the amount of times they have used each respective appliance in a day, and in turn, calculate the amount of water used. This isn’t indicated by the wireframes below but I will probably add an area above the relevant tiles where the user can say how many times they flushed for instance.

system diagram

IMG_0244

interaction diagram

IMG_0245

IMG_0246

interaction continued with software interface

IMG_0243

IMG_0241

bill of materials (per sensor)

  • Arduino Micro with Headers – 5V 16MHz
  • Liquid Flow Meter – Plastic 1/2″ NPS Threaded
  • Adafruit CC3000 WiFi Breakout with Onboard Ceramic Antenna
  • Adafruit Perma-Proto Mint Tin Size Breadboard PCB
  • 2 LEDs
  • (Possible) USB + Serial Backpack Kit with 16×2 RGB backlight negative LCD
  • (Possible) Mini Metal Speaker w/ Wires – 8 ohm 0.5W
  • Wires
  • 5V 2A (2000mA) switching power supply – UL Listed or 9V Battery (Might do one sensor with battery)
  • Voltage Regulator
  • Faucet adapter 1 1/4″ –> 1/2″
  • Faucet adapter 1/2″ –> 1 1/4″
  • http://vimeo.com/79552842 (reference for enclosures)
    • Cut Acrylic
    • 4 Standoffs

Project Plan and Timeline

Well Tap is my final project for both ICM and PCOM, and therefore, I need to plan my milestones accordingly since the ICM final is due two weeks before PCOM. With that in mind, and with the presentation date of 12.3.14, here is my project timeline below:

week 1
  • Get flowmeter sensor working on the arduino Uno
  • Get alpha version of data visualization dashboard up in p5 or processing for ICM user test (not pulling data from sensors)—Daily usage page
Week 2
  • Familiarize myself with Wifi comms between CC3000 WiFi Shield and dweet.io
  • Get 1 flowmeter sensor working on the arduino micro with CC3000 WiFi Shield. Have it communicate to dweet.io database.
  • Refine daily usage page and add weekly usage page of data visualization dashboard for final ICM presentation (no data from sensors)
Week 3
  • Get 2nd Well Tap sensor communicating with dweet.io
  • Get dweet.io communicating with Processing
  • Start preparing final fabrication
    • putting together enclosure for arduino, wifi breakout and PCB
      • soldering circuits on PCB
      • laser cut acrylic
    • putting together enclosure for flow sensor and LEDS attached to faucet
WEEK 4
  • Finalize fabrication
  • QA sensors and dashboard
  • Start tracking my water usage for demo purposes (Winter Show)

Submit a comment