Final Project: Water for Aduvan¶

Final Project Presentation¶

The final project presentation explains the motivation, system architecture, design process, implementation, testing, results, and future development of Water for Aduvan.

View / Download Final Project Presentation

Final Project Video¶

The video demonstrates the Water for Aduvan system, including the water treatment process, electronic control system, sensors, automated valves, and IoT monitoring dashboard.

Watch the Final Project Video

Project Documentation¶

The following pages contain the detailed documentation for the project:

What does it do?¶

Water for Aduvan is a smart, modular water purification and monitoring system designed to improve access to safer drinking water in underserved communities.

The system combines physical filtration, water-quality sensing, automated flow control, embedded electronics, and IoT monitoring. The system monitors water conditions and uses an ESP32-based controller to manage pumps and solenoid valves.

The current prototype focuses on filtration, turbidity monitoring, flow monitoring, automated water routing, and remote monitoring.

Future development will integrate additional treatment technologies, including electrocoagulation and photocatalytic treatment, through planned collaboration and further experimentation.

The system follows a modular architecture so that treatment, sensing, control, and monitoring components can be independently improved or replaced.

Main system functions¶

Water filtration
Turbidity monitoring
Flow-rate monitoring
Automated pump control
Automated solenoid-valve control
ESP32-based embedded control
MQTT/IoT communication
Remote dashboard monitoring
Local system status monitoring
Modular physical construction

Who has done what beforehand?¶

Water purification and water-quality monitoring systems have been developed using technologies such as physical filtration, reverse osmosis, electrocoagulation, photocatalysis, embedded systems, and IoT monitoring.

Existing systems provided useful references for the design of Water for Aduvan. However, many water-treatment systems are designed for larger installations or depend on components and infrastructure that can be difficult to maintain in rural communities.

Water for Aduvan explores how digital fabrication, embedded electronics, sensors, and IoT technologies can be combined into a lower-cost and locally maintainable system.

The project is particularly interested in:

Affordable water-treatment technologies
Local fabrication
Automated monitoring
Community-scale deployment
Modular system architecture
Real-time water-quality information
Future integration of advanced treatment technologies

What did you design?¶

I designed and developed several mechanical, electronic, software, and system-level components.

1. Sensor mounting and housing¶

Custom mounting components were designed to position and protect the sensors within the water-treatment system.

The designs were developed to make sensor installation, maintenance, and replacement easier.

2. Electronics control box¶

A custom enclosure was designed to contain the electronic control system.

The enclosure houses components including:

ESP32 controller
Power regulation components
Relay/control electronics
Sensor connections
Pump and valve control connections
Wiring and connectors

Control Box

3. Custom control PCB¶

A custom PCB was designed to organize the connections between the microcontroller, sensors, power system, and actuators.

PCB Design

PCB Layout

The PCB was fabricated and assembled as part of the electronics development process.

4. IoT dashboard¶

A dashboard was developed to visualize system information and support remote monitoring.

Dashboard

Monitoring Kiosk

5. Water-treatment system layout¶

The plumbing and treatment modules were arranged to provide a controlled path for water through the system.

System Integration

What sources did you use?¶

The project was informed by technical documentation, research, previous projects, and component documentation.

Sources included:

Fab Academy documentation
Previous Fab Academy final projects
ESP32 documentation
Sensor manufacturer datasheets
IoT and MQTT documentation
Research publications
ResearchGate
Wikipedia for preliminary background research
YouTube tutorials and technical demonstrations
ChatGPT for technical guidance, debugging assistance, explanations, and development support
WHO drinking-water guidance and related water-quality references

Technical information from online sources was compared with component datasheets and practical testing wherever possible.

What materials and components were used?¶

Mechanical components¶

PVC pipes and fittings
PPR pipe and fittings
Acrylic
Fasteners
Filtration media
Sand
Gravel/pebbles
Activated/biochar filtration media
Photocatalytic materials for future development
3D-printed components

PPR Components

Electronic components¶

ESP32 microcontroller
Turbidity sensor
Flow-rate sensor
Solenoid valve
Water pump
Relay/control components
LCD display
DC-DC buck converters
Custom PCB
Wiring and connectors
Power supply/battery components

Electronic Components

Where did the materials and components come from?¶

Electronic components were obtained from a combination of:

Fab Lab Winam inventory
Local electronics suppliers
Online marketplaces
Components sourced through project partners and international suppliers

Mechanical materials were obtained mainly from:

Local hardware stores in Kisumu
Plumbing suppliers
Electronics suppliers
Fab Lab Winam fabrication facilities

The project intentionally used materials and components that can be sourced or substituted locally where possible.

How much did it cost?¶

The estimated prototype cost was approximately KES 16,900.

Item	Estimated Cost (KES)
ESP32-S2 Mini	1,200
Turbidity Sensor	1,500
Flow Sensor	800
Solenoid Valve	1,500
Relay Module	400
PCB Materials	1,000
Aluminium Electrodes	1,500
Acrylic and Fabrication Materials	3,000
Filtration Media	2,000
Plumbing Components	2,500
Miscellaneous	1,500
Total	16,900

The cost represents the prototype and does not necessarily represent the expected cost of a future optimized production version.

What parts and systems were made?¶

Water-treatment system¶

The physical treatment system was assembled from several modular sections.

Sediment holder¶

A sediment/screening component was developed to remove larger particles before water enters the finer filtration stages.

Sediment Holder

Filtration chamber¶

A filtration chamber was constructed using locally available materials and filtration media.

Filtration Chamber

Sensor holder¶

A custom sensor holder was developed to position the water-quality sensor in the appropriate section of the system.

Sensor Holder

Electronics system¶

The electronics system provides sensing, control, power management, and communication.

Electronics System

The system includes:

ESP32 controller
Sensor inputs
Pump control
Solenoid-valve control
Power regulation
Local display
IoT communication

Software system¶

The software system consists of:

ESP32 embedded firmware
Sensor data acquisition
Sensor processing
Automated control logic
MQTT communication
IoT dashboard
System monitoring

The embedded controller receives sensor information and uses programmed logic to control the water-treatment system.

Mechanical system¶

The mechanical system includes:

Pipe connections
Filtration chambers
Sensor holders
Electronics enclosure
Mounting components
3D-printed components

Pipe Connections

PPR Connections

Enclosure

What processes were used?¶

The project used several Fab Academy processes. For each process, I documented both the process and what was made using it.

2D CAD Design → PCB and component layouts¶

2D CAD and electronic design tools were used to create:

PCB layouts
PCB traces
Board outlines
Component placement
System layout drawings

The PCB design was prepared for digital fabrication and later milled.

3D CAD Modeling → Sensor Housing and Control Box¶

3D CAD modeling was used to design:

Sensor housing
Sensor mounting fixtures
Control/electronics box
Mechanical brackets
Other custom mounting components

The CAD models allowed dimensions and component placement to be tested before fabrication.

Sensor Housing CAD¶

CAD model to be displayed here:

Sensor Housing CAD

Control Box CAD¶

CAD model to be displayed here:

Control Box CAD

3D Printing → Custom Mechanical Parts¶

3D printing was used to manufacture custom mechanical components that could not easily be produced using standard hardware.

Examples include:

Sensor holders
Sensor housing components
Mounting fixtures
Electronics enclosure components

3D Printed Control Box

PCB Design → Custom Control PCB¶

A custom PCB was designed for the Water for Aduvan control system.

The PCB was designed to provide organized connections between:

ESP32
Sensors
Pump
Solenoid valves
Power system
Control electronics

PCB Design

PCB Layout

PCB Fabrication → Milling the PCB¶

The PCB was fabricated using a digital milling process.

The fabrication workflow included:

Preparing the PCB design.
Checking the board outline and clearances.
Exporting the required fabrication files.
Preparing the milling toolpaths.
Fixing the copper board to the milling machine.
Milling the PCB traces.
Milling the board outline.
Cleaning the board.
Inspecting the milled traces.
Drilling/fabricating required holes.
Testing electrical continuity.

Milled PCB

The milling process required careful adjustment of tool settings, trace widths, clearances, and board alignment to avoid damaged traces.

PCB Assembly → Soldering and Testing¶

After milling, electronic components were soldered onto the PCB.

The assembly process included:

Cleaning the PCB
Inspecting the milled traces
Installing components
Soldering components
Checking solder joints
Testing continuity
Connecting sensors and actuators
Power testing
Testing communication with the ESP32

PCB Assembly

The completed PCB was then integrated into the control system.

Programming → ESP32 Firmware¶

The ESP32 was programmed to:

Read sensor values
Process sensor data
Control pumps
Control solenoid valves
Monitor system status
Send data through MQTT
Support IoT dashboard monitoring

The firmware was tested incrementally during system integration.

IoT Development → Remote Monitoring Dashboard¶

An IoT dashboard was developed to display information from the water-treatment system.

The dashboard provides a way to monitor system information remotely and visualize sensor data.

IoT Dashboard

Plumbing and System Integration → Complete Water System¶

The individual mechanical, electronic, and software subsystems were integrated into one working prototype.

The integration included:

Connecting filtration stages
Installing sensors
Connecting the pump
Installing solenoid valves
Connecting the control electronics
Connecting the power system
Installing the custom PCB
Connecting the IoT system
Testing the complete water path

Integrated System

Sensor Calibration → Water Quality Monitoring¶

The turbidity and flow sensors were calibrated and tested before being used in the automated control system.

The calibration process involved comparing sensor readings under different operating conditions and adjusting the software interpretation of the readings.

What questions were answered?¶

The project investigated the following questions:

Can a low-cost ESP32-based controller manage a water-treatment process?
Can turbidity measurements be used for automated water-flow decisions?
Can water-quality information be monitored remotely?
Can digital fabrication be used to produce custom water-treatment components?
Can locally available materials be incorporated into a modular water-treatment system?
Can the system architecture be replicated for community-scale applications?
What limitations need to be addressed before community deployment?

What worked?¶

The following parts of the prototype worked successfully:

Turbidity monitoring provided useful water-quality feedback.
The ESP32 successfully processed sensor information.
Pump control worked.
Solenoid-valve control worked.
MQTT communication successfully transmitted system information.
Dashboard visualization worked.
The custom PCB was fabricated and assembled.
The mechanical and electronic subsystems could be integrated.
The modular architecture allowed individual components to be tested separately.

What did not work or needs improvement?¶

Several challenges were identified during development.

Turbidity calibration¶

The turbidity sensor required repeated calibration and testing to obtain consistent readings.

Flow measurement¶

Flow measurements were less stable at low flow rates and require further calibration.

Filtration speed¶

The filtration process was slower than desired. Future designs will investigate improved filtration media, flow paths, pump selection, and treatment stages.

Advanced treatment stages¶

Electrocoagulation and photocatalytic treatment are part of the future development roadmap and were not fully implemented in the current prototype.

How was it evaluated?¶

The prototype was evaluated through functional testing of the electronic, mechanical, software, and water-treatment subsystems.

Evaluation Metric	Result
Water flow control	Pass
Sensor operation	Pass
Data transmission	Pass
Dashboard visualization	Pass
Turbidity monitoring	Pass
Automated decision making	Pass
Continuous system operation	Pass
Mechanical integration	Pass
PCB operation	Pass

Water samples were compared before and after filtration to evaluate changes in turbidity.

Further laboratory water-quality testing is required before making claims about drinking-water safety or community deployment.

What are the implications?¶

Water for Aduvan demonstrates how digital fabrication, embedded electronics, sensing, automation, and IoT can be combined to develop locally manufacturable water-treatment technologies.

The project has potential implications for:

Affordable water-treatment systems
Community-level monitoring
Local manufacturing
Technical skills development
Real-time water-quality monitoring
Modular treatment systems
Rural and peri-urban water infrastructure

The project is aligned with UN Sustainable Development Goal 6: Clean Water and Sanitation.

Partnership and Future Development¶

Future development will focus on improving the treatment process and preparing the system for real-world testing.

Planned developments include:

Integration of an electrocoagulation chamber
Photocatalytic treatment
Partnership with Toyota for further development
Additional water-quality sensors
TDS/conductivity monitoring
Improved turbidity sensing
Solar-powered operation
Battery-powered portable operation
Improved filtration speed
Improved mechanical enclosure
More robust PCB design
OTA firmware updates
Community pilot testing
Deployment around communities in the Lake Victoria region

The future objective is to transform the prototype into a more robust, portable, energy-efficient, and maintainable water-treatment platform.

Business Model

Use of AI and ChatGPT¶

Artificial intelligence tools, including ChatGPT, were used during development as a supporting tool.

ChatGPT was not treated as the author of the project. It was used for activities such as:

Explaining technical concepts
Troubleshooting programming errors
Reviewing code logic
Suggesting debugging approaches
Explaining sensor behaviour
Assisting with documentation structure
Improving technical writing
Generating initial code examples where applicable

Any code that was not originally written by me is identified as AI-assisted or sourced from the relevant reference.

ChatGPT prompts used¶

Examples of prompts used during development include:

“Help me write ESP32 code to read a turbidity sensor and control a solenoid valve based on the sensor reading.”

“Help me troubleshoot why the ESP32 ADC is reading 4095 from the turbidity sensor.”

“Explain how to connect an ESP32 to a turbidity sensor and flow sensor.”

“Help me debug the MQTT communication between the ESP32 and the IoT dashboard.”

“Explain how I can calibrate the turbidity sensor.”

“Help me improve this ESP32 code and explain the changes.”

“Help me document my Fab Academy final project.”

Where AI-generated code was incorporated into the project, the relevant project page should identify the code as ChatGPT-assisted and include the actual prompt used.

Design Files¶

All original design files used to fabricate the Water for Aduvan prototype should be provided below.

CAD Files¶

PCB Design Files¶

Manufacturing PCB Files¶

text

3D Printing Files¶

text

Source Code¶

Project Summary¶

Water for Aduvan combines water treatment, digital fabrication, electronics, embedded programming, sensing, automation, and IoT monitoring into one integrated prototype.

The project demonstrates the complete process from problem identification → concept development → CAD design → electronic design → PCB fabrication → programming → mechanical fabrication → system integration → testing → future development.

The project will continue beyond Fab Academy through further treatment development, partnership, testing, and potential community deployment.

License¶

Documentation and Creative Work¶

Unless otherwise stated, the original documentation, photographs, diagrams, CAD designs, PCB designs, illustrations, and other creative materials created for the Water for Aduvan project by Charles Otieno Wangara are licensed under the:

Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0)

This means that others may share and adapt this work for non-commercial purposes, provided that they:

Give appropriate credit to Charles Otieno Wangara and the Water for Aduvan project.
Provide a link to the CC BY-NC-SA 4.0 license.
Indicate if changes were made.
Distribute adaptations under the same license.

View the full CC BY-NC-SA 4.0 license

Software¶

The source code developed for the Water for Aduvan project is released under the MIT License, unless otherwise stated.

This includes original software such as:

ESP32 firmware
Sensor-control software
MQTT communication
IoT dashboard
Backend services
Frontend application

The MIT License permits others to use, copy, modify, merge, publish, distribute, sublicense, and sell copies of the software, provided that the original copyright and license notice are retained.

The software license does not change the license of the project documentation, CAD designs, photographs, or other creative materials.

Third-Party Materials¶

Third-party materials used or referenced in this project, including manufacturer datasheets, logos, external images, libraries, research papers, and other copyrighted materials, remain subject to their respective licenses and are not automatically covered by the licenses above.