Team DAMP

Use this template to document your progress! (Add notes anywhere you deem necessary)

We are DAMP+J, a team formed during the hackathon to explore how IoT can be used to influence individual behavior for sustainability and well-being. Our group brings together complementary skills in embedded systems, software engineering, UX design, and environmental awareness.

Our approach prioritizes simplicity, functionality, and measurable impact. Rather than building complex systems for their own sake, we focused on applying accessible technologies to address everyday habits that have long-term environmental consequences.

Our final project, FRoast, reflects this mindset—a behavior-aware smart fridge prototype designed to promote conscious food habits, reduce waste, and make sustainability both personal and actionable.

- Jorge Leon - Marcia Rodrigues - Daniella Fajardo - Prachi Singhal - Abdulkadir Abubakar

We began the ideation process by exploring IoT-based solutions that focus on changing individual behavior to promote sustainability, accessibility, and well-being. We explored several IoT-based concepts that aim to nudge individuals toward sustainable, healthy, and inclusive behaviors, using a combination of sensors, actuators, for IoT tools. Below are the documented ideas from our brainstorming process:

• Goal: Reduce water and energy waste by encouraging shorter, more efficient showers.
• What it’s about: This system monitors shower duration using steam and temperature sensors. It gives real-time feedback using buzzers, RGB lights, or LCD messages when a user exceeds optimal limits. The aim is to build awareness of water usage and promote sustainable bathing habits.
• Behavioral Nudge: Timed alerts and eco-scores after each shower encourage users to reduce consumption over time.
• Uses:
  • Steam and temperature sensors to track shower duration and frequency.
  • Real-time nudges (via buzzer, RGB lights, or LCD display) when limits are exceeded.
  • Eco-feedback or daily stats to reward water-efficient behavior.

• Goal: Help individuals become more energy-conscious in their personal spaces.
• What it’s about: This system acts as a digital coach inside a user’s room. It monitors air conditioning use, artificial lighting, fan activity, and motion detection to identify unnecessary resource consumption. Feedback is delivered via lights, screen messages, and audio cues.
• Behavioral Nudge: Personalized reminders like “Natural light available” or “Fan use unnecessary” reinforce energy-saving habits without being intrusive.
• Uses:
  • Air Conditioning Use: Warns if temperature is unnecessarily low/high (Temp & Humidity Sensor).
  • Lighting Habits: Detects unnecessary artificial light use when natural light is sufficient (Light Sensor).
  • Fan Usage: Suggests natural ventilation over active cooling (Fan + Temp Sensor).
  • Idle Energy Consumption: Alerts if no motion is detected but devices/lights are left on.

• Goal: Encourage more sustainable dietary choices by reducing unnecessary meat consumption.
• What it’s about: While not fully developed, the concept involved using fridge/cooking access data to infer meat-heavy consumption patterns. The system would then prompt users with healthy plant-based alternatives or reminders through the display or audio.
• Behavioral Nudge: Frequency-based alerts or daily summaries showing excessive food access could lead to more mindful eating choices.
• Uses:
  • Logs fridge/cooking access times and compares with diet goals.
  • RFID integration could track user habits and trigger alternative food suggestions.
  • While not implemented, this idea is aligned with the final fridge project selected.

• Goal: Combine energy efficiency with personalized accessibility for users with diverse needs.
• What it’s about: EcoAccess Station uses RFID to identify users and adapt room conditions based on their physical or sensory requirements (e.g., mobility-impaired, heat-sensitive). It combines motion, temperature, and humidity sensors to optimize energy use while respecting user comfort.
• Behavioral Nudge: Shows how responsive systems can balance inclusion with sustainability, teaching users to be more conscious about shared resource usage.
• Uses:
  • Adjusts lighting, fan speed, or notifications based on user profile.
  • Reduces energy waste by activating systems only when needed.
  • Uses motion, temperature, and humidity sensors to evaluate environmental needs.
  • Provides accessible feedback through color-coded LEDs and simple audio/text messages.

• Goal: Promote healthier eating habits and reduce unconscious fridge use through personalized feedback.
• What it’s about: Coach Fridge uses RFID to identify users and monitors their frequency, duration, and timing of fridge access using motion, steam, gas, and temperature sensors. Based on the behavior, it provides real-time nudges such as reminders to avoid late-night snacking or to shorten door-open time.
• Behavioral Nudge: Uses lights, sound, and friendly screen messages to mimic a digital coach that helps users make better eating decisions over time.
• System Features:
  • Identifies users with RFID to personalize the experience and track individual habits.
  • Monitors fridge usage patterns, including:
    • How often the fridge is accessed (access frequency)
    • What time of day it’s used (timing)
    • How long the door stays open (duration)
  • Uses sensor data to detect and respond to specific behaviors such as:
    • Late-night snacking — alerts the user if they frequently access the fridge during sleep hours
    • Leaving the fridge open too long — uses temperature and motion sensors to trigger reminders
    • Storing hot food — detects high humidity or steam and warns about food safety and energy waste
    • Poor food rotation — infers habits from repetitive access patterns without variation
    • Not eating enough fruits — tracks lack of food diversity through access frequency and meal prep signals
    • Ignoring seasonal energy needs — adapts feedback based on external temperature or predefined seasonal settings
  • Provides real-time feedback through:
    • LCD messages with health or sustainability tips
    • RGB/LED lights to indicate usage behavior (e.g., green = ideal, red = needs improvement)
    • Buzzer sounds for soft alerts and behavior warnings
  • Acts like an AI-powered fridge coach, using rule-based logic and behavior history to offer supportive, timely nudges and guide better food habits over time.

measuring-user-behavior-change-in-froast.pdf

The project is a smart fridge system designed to monitor and influence users’ eating habits through real-time feedback and personalized behavior tracking. It uses RFID identification, sensors, and actuators to encourage healthier and more sustainable routines such as avoiding late-night snacking, reducing energy waste, and improving food management.

Each time a user approaches the fridge, they are identified using an RFID tag. The system monitors fridge access patterns — how often, when, and for how long the fridge is used — and cross-references this with sensor data (e.g., steam, gas, temperature, motion) to detect potential habits like:

• Frequent late-night visits
• Leaving the fridge door open for too long
• Storing hot food immediately
• Poor food rotation or repetitive access
• Lack of healthy or diverse food preparation

Real-time nudges are provided through LEDs, buzzer sounds, and messages on an LCD display, mimicking the role of a digital “fridge coach.”

• User Identification: RFID module links behavior to individual users.
• Usage Tracking: Logs frequency, timing, and duration of fridge access.
• Behavior Detection:
  • Steam/Gas sensors to detect hot food or cooking activity
  • Temperature sensor to detect door-open duration
  • Motion sensor to measure presence near the fridge
• Feedback Mechanisms:
  • LCD display for personalized messages and tips
  • RGB/LED lights to indicate behavior status (e.g., green = healthy, red = excessive)
  • Buzzer for alerts when behaviors exceed limits
• Habit Reinforcement:
  • Encourages mindful eating
  • Reduces unnecessary energy consumption
  • Simulates an AI-powered coach using rule-based logic

While the current prototype of FRoast successfully demonstrates real-time behavior tracking and feedback mechanisms, several enhancements could significantly broaden its impact, adaptability, and long-term user engagement. The following proposals aim to transform the system from a fixed, rule-based framework into a dynamic, user-centered behavioral assistant.

Integrating basic food identification features—such as barcode scanning, NFC tags, or weight sensors—could enable FRoast to provide context-specific feedback based on the type of food being stored. This would allow the system to:

• Encourage the storage of healthier food options
• Discourage excessive processed food consumption
• Track diversity in food selection

By associating RFID tags with specific food items or enabling manual item logging, FRoast could offer inventory tracking capabilities. This would support:

• Food rotation recommendations
• Alerts for items nearing expiration
• Waste reduction via proactive usage reminders

To enhance accessibility, especially for users with visual impairments or limited literacy, the system could include auditory feedback through basic speech synthesis. Voice guidance could supplement or replace LCD text, making the system more inclusive for children, elderly users, or individuals with disabilities.

Incorporating ambient temperature sensing would enable FRoast to adjust its behavioral prompts based on seasonal energy considerations. For example:

• Stricter alerts during hot weather when open-door cooling loss is more severe
• Relaxed thresholds during colder months when external temperatures mitigate loss

Introducing a lightweight machine learning layer could allow the system to evolve beyond fixed thresholds. Using historical behavior data, the model could:

• Adapt the intensity and tone of feedback to each user’s responsiveness
• Identify subtle habit shifts and adjust nudging strategies accordingly
• Differentiate between habitual and intentional exceptions

FRoast is designed as more than a sensor-based monitoring system—it is a behavior-oriented intervention tool focused on reducing food and energy waste in domestic environments. Through RFID-linked user profiles, sensor feedback, and context-aware prompts, the system promotes conscious engagement with everyday habits related to food storage and refrigerator use.

Refrigerators contribute significantly to household energy consumption and are often involved in avoidable food waste due to poor usage habits. Traditional tracking alone provides data but fails to initiate meaningful change. FRoast addresses this gap by integrating real-time interventions—turning passive monitoring into active behavioral guidance.

The premise is simple but effective: raising user awareness through timely, personalized prompts can influence decision-making at the moment it matters most. By embedding behavioral nudges into daily routines, the system seeks to support more sustainable and health-conscious food practices over time.

Each user is identified via RFID, enabling individualized tracking and tailored feedback. The following behaviors are monitored:

• Fridge Door Access Patterns
  • Frequency of openings per day
  • Duration of door open per session

• Nighttime Usage
  • Identification of fridge use during designated rest hours

• Steam Events
  • Detection of hot food insertion through steam sensors, indicating inefficient storage behavior

• Gas Events
  • Detection of spoilage or strong odors, signaling delayed disposal or neglected items

• Fruit and Compartment Use
  • Patterns of compartment interaction, used as proxies for dietary diversity or habit stagnation

To assess the effectiveness of FRoast in modifying behavior, both quantitative and qualitative metrics are collected:

Quantitative Indicators

• Average duration of fridge door openings over time
• Number of late-night access events per week
• Frequency of steam or gas sensor activations
• Interaction rates with the fruit compartment or healthier food zones
• Frequency of sustainability warnings recorded via the internal log or connected web interface

Qualitative Indicators

• User-reported changes in snacking frequency or food discipline
• User preference tracking based on response to feedback tone (e.g., comedic, neutral, advisory)

The long-term objective of FRoast is to instill sustainable behavior without requiring significant user effort or intrusive systems. By embedding behaviorally-informed micro-interventions into an everyday appliance, the system can contribute to:

• Improved food storage habits and reduced spoilage
• Lower average energy usage through better door discipline
• Reduced impulsive eating, particularly at night
• Increased user engagement with sustainability as a personal responsibility

Through repeated exposure to feedback and subtle reinforcement, FRoast supports gradual habit reformation, enhancing both environmental outcomes and personal well-being.

impactxchange-hackathon-3.pdf

import uasyncio as asyncio
from machine import Pin, PWM, SoftI2C, ADC, DAC
from i2c_lcd import I2cLcd
import time
import dht
from mfrc522_i2c import mfrc522
import collections
 
class MessageQueue:
    def __init__(self):
        self.queue = []
 
    async def put(self, item):
        self.queue.append(item)
 
    async def get(self):
        while not self.queue:
            await asyncio.sleep(0.1)  # Non-blocking wait
        return self.queue.pop(0)
 
message_queue = MessageQueue()
 
# Constants
SERVO_PIN = 5
BUZZER_PIN = 25
I2C_SCL = 22
I2C_SDA = 21
I2C_ADDR = 0x27
RFID_ADDR = 0x28
AUTHORIZED_ID = 547
SERVO_OPEN = 125
SERVO_CLOSE = 28
BUZZER_FREQ1 = 1000
BUZZER_FREQ2 = 1500
ALARM_DURATION = 0.5
HUMIDITY_THRESHOLD = 30
TEMP_THRESHOLD = 27
GAS_PIN = 23
GAS_SENSOR = Pin(GAS_PIN, Pin.IN, Pin.PULL_UP)
GAS_THRESHOLD = 70  # Adjust based on sensor calibration
 
# State
door_opened = False
open_time = 0
 
# I2C + LCD setup
i2c = SoftI2C(scl=Pin(I2C_SCL, Pin.OUT, pull=Pin.PULL_UP),
              sda=Pin(I2C_SDA, Pin.OUT, pull=Pin.PULL_UP),
              freq=100000)
lcd = I2cLcd(i2c, I2C_ADDR, 2, 16)
 
# Buzzer setup
buzzer = PWM(Pin(BUZZER_PIN))
buzzer.duty(0)
 
# RFID setup
rc522 = mfrc522(I2C_SCL, I2C_SDA, RFID_ADDR)
rc522.PCD_Init()
rc522.ShowReaderDetails()
 
# DHT Sensor setup
DHT_SENSOR = dht.DHT11(Pin(17))
# === STEAM SETUP ===
STEAM_SENSOR = ADC(Pin(34))  # Analog pin (VP/A2)
STEAM_SENSOR.atten(ADC.ATTN_11DB)  # Full voltage range (0-3.3V)
STEAM_THRESHOLD = 800  # Analog value above this = steam detected
 
current_user = "Unknown"
 
 
async def move_servo(position):
    servo = PWM(Pin(SERVO_PIN))
    servo.freq(50)
    servo.duty(position)
    await asyncio.sleep(0.5)
    servo.deinit()
 
async def open_door():
    global door_opened, open_time
    print("Opening door")
    await move_servo(SERVO_OPEN)
    open_time = time.time()
    door_opened = True
 
async def close_door():
    global door_opened
    print("Closing door")
    await move_servo(SERVO_CLOSE)
    door_opened = False
 
 
async def play_alarm():
    print("Playing alarm")
    buzzer.duty(512)
    end_time = time.ticks_ms() + 2000
    while time.ticks_ms() < end_time:
        buzzer.freq(BUZZER_FREQ1)
        await asyncio.sleep(ALARM_DURATION)
        buzzer.freq(BUZZER_FREQ2)
        await asyncio.sleep(ALARM_DURATION)
    buzzer.duty(0)
 
 
async def scroll_text(text, row, delay=0.5):
    if len(text) <= 16:
        lcd.move_to(0, row)
        lcd.putstr(text + ' ' * (16 - len(text)))
        return
    for i in range(len(text) - 15):
        lcd.move_to(0, row)
        lcd.putstr(text[i:i + 16])
        await asyncio.sleep(delay)
 
async def message_handler():
    while True:
        upper, lower = await message_queue.get()
        lcd.clear()
        await scroll_text(upper, 0)
        await scroll_text(lower, 1)
 
# --- ASYNC TASKS ---
 
async def monitor_rfid():
    global door_opened, open_time
    while True:
        buzzer.duty(0)
        if rc522.PICC_IsNewCardPresent():
            if rc522.PICC_ReadCardSerial():
                data = sum(rc522.uid.uidByte[0:rc522.uid.size])
                print("Card UID:", data)
                current_user = data
                print("USER,TYPE={}, TIMESTAMP={}".format(current_user, time.time()))
 
                if data == AUTHORIZED_ID:
                    if door_opened:
                        print("LOG,TYPE=OPEN_DOOR, TIMESTAMP={}".format(time.time()))
                        duration = time.time() - open_time
                        print("Duration:", duration)
                        await close_door()
                        if duration > 15:
                            await play_alarm()
                            await message_queue.put(("Your indecision is heating the planet.", ""))
                            print("WARNING,TYPE=OPEN_DOOR, TIMESTAMP={}".format(time.time()))
 
                        else:
                            await message_queue.put(("Thanks", "For closing me"))
                    else:
                        await open_door()
 
                        # Time-based snacking roast
                        current_hour = time.localtime()[3]
                        if current_hour >= 23 or current_hour < 5:
                            await message_queue.put(("You hungry or just emotionally empty again?", "Go to sleep, goblin."))
                            print("WARNING,TYPE=LATE_NIGHT, TIMESTAMP={}".format(time.time()))
                        else:
                            await message_queue.put(("Welcome!", "Door is open"))
 
                else:
                    print("Not authorized")
                    await message_queue.put(("Access Denied", "Invalid Card"))
        await asyncio.sleep(1)
 
def monitor_steam_sensor():
    print("🔍 Steam sensor activated...")
    while True:
        try:
            analog_value = STEAM_SENSOR.read()
            print("Steam sensor value:", analog_value)
 
            if analog_value > STEAM_THRESHOLD:
                await message_queue.put(("The fridge is not your sauna, pal.", ""))
                print("WARNING,TYPE=HOT_FOOD, TIMESTAMP={}".format(time.time()))
 
            await asyncio.sleep(2)
 
        except Exception as e:
            print("DHT read error:", e)
        await asyncio.sleep(2)
 
async def monitor_humidity():
    while True:
        try:
            DHT_SENSOR.measure() # Start DHT11 to measure data once.
            # Call the built-in function of DHT to obtain temperature
            # and humidity data and print them in “Shell”.
            temperature_value = DHT_SENSOR.temperature()
            humidity_value = DHT_SENSOR.humidity()
            print('temperature:', temperature_value,'℃','humidity:', humidity_value,'%')
            if temperature_value > TEMP_THRESHOLD:
                await message_queue.put(("Want to cool the Sahara too?", ""))
                print("WARNING,TYPE=HOT_FOOD, TIMESTAMP={}".format(time.time()))
            elif humidity_value > HUMIDITY_THRESHOLD:
                await message_queue.put(("Too hot in here!", "Not a sauna, pal."))
                print("WARNING,TYPE=HOT_FOOD, TIMESTAMP={}".format(time.time()))
        except Exception as e:
            print("DHT read error:", e)
        await asyncio.sleep(2)
 
async def monitor_compartments():
    button1 = Pin(16, Pin.IN, Pin.PULL_UP)
 
    button2 = Pin(27, Pin.IN, Pin.PULL_UP)                          #close
 
    #Nest two for loops to make the module repeatedly display five states of red, green, blue, white and OFF.
 
    while True:
        try:
            btnVal1 = button1.value()  # Reads the value of button 1
 
            #print("button1 =",btnVal1)  #Print it out in the shell
 
            if(btnVal1 == 0):
                await message_queue.put(("Those vitamins won’t drink themselves.", ""))
 
                print("LOG,TYPE=VITAMIN, TIMESTAMP={}".format(time.time()))
 
            btnVal2 = button2.value()
 
            if(btnVal2 == 0):
                await message_queue.put(("Food compartment opened.", ""))
 
                print("LOG,TYPE=FOOD, TIMESTAMP={}".format(time.time()))
        except Exception as e:
            print("Compartments read error:", e)
        await asyncio.sleep(0.5)
 
async def monitor_gas():
    while True:
        try:
            gas_level = GAS_SENSOR.value()
            print("Gas Level:", gas_level)
            if not gas_level:
                await message_queue.put(("You really like cultivating diseases, huh?", "Did your fruit die?"))
                print("WARNING,TYPE=FOOD_SPOILAGE, TIMESTAMP={}".format(time.time()))
        except Exception as e:
            print("Gas sensor error:", e)
        await asyncio.sleep(3)
 
# --- MAIN ---
async def main():
    await asyncio.gather(
        monitor_rfid(),
        monitor_humidity(),
        monitor_compartments(),
        monitor_steam_sensor(),
        monitor_gas(),
        message_handler()
    )
 
asyncio.run(main())