The Laundry Basket Wants You to Start the Washing Machine

The Laundry Basket Wants You to Start the Washing Machine!

Tired of laundry day hassles?

Meet the Smart Laundry Basket System – your ultimate laundry assistant!

This revolutionary basket does more than hold your clothes. It tracks and classifies each item, ensuring perfect washes every time.

How does it work?

Simply toss in your garments. The system uses cutting-edge AI to identify fabric, color, and temperature needs, then builds the ideal washing list.

When it’s ready, you get an alert to start the machine with the perfect settings.

Say goodbye to sorting and overloads.

Embrace the future of laundry and say hello to effortless laundry days and hello to more time for the things you love.

For my instructable, I've gathered up the following things:

• my laptop
• Raspberry Pi 5
• Freenove Projects kit for Raspberry Pi
• SD Card 32 Gb
• a laundry basket
• cooling for the Raspberry Pi
• webcam
• cables

The PIR detects a person coming to the laundry basket and records a video of the content of the laundry basket.

Preparing laundry is like crafting a perfect recipe.

You start by sorting whites from colors and organizing by fabric type.

• Whites get a 30°C wash with a whitener to keep them bright.
• Light-colored cottons go in at 30°C,
• Both get 60°C when they need a deeper clean.
• Synthetics, wool and delicates require a gentle 30°C cycle,

Each step ensures your clothes are cleaned with care and precision.

Define a classification for the AI model.

• color 30 degrees
• color 60 degrees
• white 30 degrees
• white 60 degrees
• wool & delicates 30 degrees

source washing program symbols

The AI model uses supervised learning and needs photos of the laundry.

Taking photos of the clothes and towels and other garments you throw in the laundry basket. Take several photos of the garment, in different positions.

Make a Roboflow account, upload your photos and start annotating!

Use bounding boxes to indicate where your laundry is and to what class it belongs.

Your model needs this to train well.

Choose the object detection model, augment your data and train your model.

https://blog.roboflow.com/getting-started-with-roboflow/

Roboflow has a lot of annotated and pretrained datasets to use too.

I used my own photos but also these:

https://universe.roboflow.com/sockensortierer3-nfayd/sockensortierer-syjwg

https://universe.roboflow.com/chris/laundry

Retrain your model to generate the best model and use metrics to validate. The confusion matrix is the best.

The Raspberry Pi needs to detect when something is added to the laundry basket.

In other words when someone approaches the laundry basket, the content of the basket has to be analysed.

The Raspberry Pi detects motion and sends a message to the laptop to take a photo of the content of the laundry basket.

To detect motion, I am using the PIR sensor on the freenove board. Adjust the time between two detections (approx 10 seconds) and the distance has to be at its minimum, which is 3 meter.

The code is running on the Raspberry Pi and communicates with the laptop with an ethernet cable. Of course you can use a Bluetooth connection too.

I like the explanations of Paul McWhorter about the PIR motion sensor a lot, as it is explained for total beginners! Thank you Paul, you helped me a lot.

Code for the Raspberry

When the laptop receives the message from the PIR sensor, it takes a photo of the content of the laundry basket.

After the photo is taken, the supervised learning model starts to classify the laundry.

A csv file is created and when a washing program has 10 garments, the laptop sends a message to the LCD display to show the program to start.

Of course the amount of garments can be defined by washing program, depending on the garments and on the washing machine.

Code for the laptop

Finally I made a simple cardboard box. It contains:

• camera
• motion sensor
• LCD display

The camera is directly connected to the laptop.

The motion sensor and LCD display to the Raspberry Pi.

The LCD display shows a message when a photo is taken, by saying "look in basket" for example

When there is enough laundry to start the washing machine. The display shows which washing program can be started.