Fresh water is becoming increasingly scarce. Between climate change, population growth, and the industrial sector, projections indicate that by 2025, over half the world’s population will live in water-stressed regions. This disparity in access to water resources is only compounded by the Western world’s excessive overconsumption habits. In 2022 alone, Australian households consumed roughly 1,779 billion litres of water – highlighting an urgent need to rethink our water management practices.
Our project, AquaSense, aims to address this issue – leveraging innovative physical computing technologies to change the way everyday Australians view water consumption. By pairing a tangible DIY system, with a user-centric, network-attached dashboard – our smart water monitor seeks to promote sustainable water practices within Australian households, simply and elegantly. In doing so, we hope to encourage individuals to take a more proactive approach towards reducing their water wastage, contributing to a more water-conscious future for generations to come.
Initial Research and Ideation
From the outset of this project, our group was excited by the idea of creating a network-attached device that could assist in managing global issues. We were inspired by projects such as WearOptimo’s Microwearable and Ceglinski & Turton’s Seabin – both of which highlight the significant impact seemingly simple technologies can have on the lives of many. Similarly, we were also motivated by the ongoing discussions about the various ‘wicked problems’ plaguing Australia – such as the recent cost of living crisis, global overconsumption, and extreme weather events.
We concluded that by incentivising ordinary Australians to reduce their resource usage, we may be able to tackle many of these issues simultaneously. Understanding this, we researched a variety of resource-related focus areas, including the electrical and gas sectors, car usage and general wastage before ultimately deciding to target water consumption.
While areas like residential trash management and electrical monitoring already had established, realised solutions (such as Numi’s Smart Bin and AGL’s recently announced Peak Energy Rewards program) – water consumption within the residential space was seemingly a gap in the market. This intrigued us, and after discussing it with collaborators, we decided on our project’s goal :
To develop a DIY, user-friendly device that promotes sustainable water management practices in residential Australian homes.
Project Development
A critical aspect of any design process revolves around rapid prototyping and iteration based on regular user testing and feedback. Our project was by no means linear and, thanks to both internal and external feedback, was able to deliver a prototype that considered a wide variety of perspectives and achieved our overall goal.
Week 7 saw our project pitched to the class. This proved invaluable, as we received many constructive points regarding potential oversights and incorrect assumptions we had made. The most prominent piece of feedback regarded the importance of specifying that our system was a DIY kit, for self-installation – necessitating the inclusion of an accompanying instruction manual. Additionally, instead of incentivising users with gamification strategies, we were advised to explore government initiatives – ultimately leading to our integration of the Australian Government’s Target 150 campaign. Further feedback suggested considering users in remote locations, living off the grid – leading us to incorporate an optional ‘tank water limit’ within our dashboard. We also refined our mechanical design further to focus specifically on garden taps, acknowledging the challenges of creating a one-size-fits-all solution and allowing for future iterations to consider other water accessory options.
Development of the Arduino
The first step in our project was the development of the physical circuitry. Multiple flow rate monitors were ordered and tested, before settling on the YF-S201 – due to its ability to fit onto taps without being overly large.
From here, existing algorithms were trialled and implemented into our sketch to calculate the literage being used, before integrating this with the web functionality. While we had originally planned to host the website directly on the Arduino itself (similarly to Zac’s ‘smart doorbell’ approach) this proved difficult, as the Arduino did not have the resources to store values for long periods.
Our original flow rate sketch (not network attached) (.ino)
Ultimately, this resulted in the decision to connect the circuit to an external server. After much experimentation with writing directly to an SQL database, we decided that parsing the values through a URL to a PHP endpoint and writing server-side provided a much more reliable solution.
Sketch used to test PHP value parsing (without attaching Flow Sensor) (.ino)
Finally, as the project progressed, various quality-of-life components were added to the physical circuitry. A longer, braided cable was introduced, to ensure the circuits were kept away from the water, and a 9v battery to power the Arduino.
Development of the Web Dashboard
A large component of our project was the development of the attached web dashboard, which served as the primary interface for our monitoring solution. Initially, the design was quite basic, simply logging data into the database and displaying it on the frontend. Once we
confirmed it was working in tandem with the circuit, we then expanded its capabilities. We began by designing low-fidelity wireframes in XD, drawing inspiration from the UI of existing data-driven dashboards. These included Ubiquity’s UniFi dashboard (used for monitoring device internet usage) and Amber Electrics’ solar panel/battery monitoring app.
Video of our Initial Web Environment
Original Low-Fidelity Wireframes
We then implemented these designs into the site placing focus on the dynamism of its data. AJAX calls were written to fetch relevant data from the database, without the need to reload – with the Chart.js library being used to display it dynamically.
After it was created, user testing highlighted the necessity of assigning a device and its data to a specific account. This led to the creation of a login portal and an overhaul of the database schema to support user-specific data management.
Finally, our Week 11 testing revealed that the UI was not inherently intuitive for all users. To address this, we introduced an onboarding process, providing relevant context and allowing users to set preferences to personalise their dashboards. It was here we decided to integrate the findings of the Target 150 initiative, using their 150L per person, per day measurement as a baseline to offer our rewards and incentives.
Development of the 3D Models
The primary components of the mechanical design included the electronic housing and the threaded connections. The threads, in particular, needed to both securely attach the water sensor to a garden tap, and allow existing tap appliances to be fitted onto the sensor.
Our models were first designed in Fusion360, taking inspiration from existing commercial plumbing solutions (including Moen’s FLO smart shutoff valve) and making use of the standardised NSW tap size. Some measurement changes were then implemented to ensure the
Arduino sat flush within the housing, before our final print. As we would later discover, the PLA used to print was permeable – a material property not ideal for this context. As such, we pivoted, swapping these for more robust, water-proof polypropylene threads. Velcro was added to the housing, allowing it to be stuck to an adjacent wall for user-centric installation.
Development of the Packaging
In line with our product’ DIY focus, we decided to create packaging for the AquaSense to help highlight the full user journey. We began by establishing some brand guidelines, based on our existing UI, before writing content for the instruction manual and accompanying box. We then designed a box to house the components – being inspired by the sleek and simple aesthetics of Google’s ‘Made By’ series. The included instruction manual was also designed and printed, undergoing various iterations as our project’s onboarding process changed. Finally, the box was lined with low-density polyethylene and styrofoam to ensure it was waterproof and snug. Our Brand Guidelines (.pdf)
Week 11 Prototyping
Week 11 User Testing
Week 11 provided us with the opportunity to demo the prototype, allowing us to explore how users interacted with our product. This was invaluable, again providing key considerations that shaped our final prototype. Some key takeaways addressed in our final product included :
- Highlighting that PLA would be prone to leakage and shrinkage.
- Suggestions that the minute view was impractical, and should rather feature an hourly overview.
- Various usability and interface suggestions regarding the web dashboard.
- Integration of the per appliance account linking and integration.
- Providing the ability for one to download user data.
Future Development
While we are happy with the end prototype and its capabilities, given more time, there are several enhancements we would make to the project before ‘putting it to market’.
Most majorly, is the WiFi onboarding. Currently, the WiFi credentials are hardcoded into the Arduino, however, we need a way for users to log in and set these values during the initial setup. While we did look at potential Arduino libraries, such as WifiManager, to solve these issues – none provided the clean, user-friendly experience, that was central to this product.
Similarly, while our current prototypes work on outdoor taps, higher-pressure appliances like sinks and washing machines need larger sensors and different connections. This necessitates offering differently-sized models of the AquaSense for specific appliances.
These are just a few of the many changes that could improve the product. Design aesthetics, web responsiveness/a mobile application, improved physical durability and more are all key areas to explore, given more time.
Critical Reflection on the Prototype
The primary goal of our project was to encourage conscious water management for residential use, and we believe this objective was mostly achieved. While this issue cannot be solved overnight and our system does have its limitations, we believe it offers a foundational step in addressing a much larger problem. The prototype itself does do a fantastic job of monitoring water usage – providing tangible usage data that individuals can use to make better informed water decisions.
Similarly, compared to other systems, our prototype offers a unique approach not yet considered by mainstream providers. Unlike IoT solutions that only target a single water tank, those designed for commercial applications or even those designed for large-scale use across apartment complexes, our system is tailored for simpler, residential markets. On top of this all, if fully realised, we believe our system would offer a more affordable, DIY approach that is integrated with government incentives – a catch-all not seen in any current consumer solutions.
Conclusion
Overall, while our project was never going to be a definitive, end-all solution to the current water crisis, we believe it does provide a practical and impactful tool for promoting water conservation at the residential level. Not only this, but we believe AquaSense effectively contributes to a growing movement to address environmental challenges through innovative technologies and user-centred design. By harnessing the power of physical computing, our project not only empowers individuals to participate in a more water-conscious future, but to consider the significant potential technology holds in addressing global societal issues.
