Alliance International Conference on Artificial Intelligence and Machine Learning (AICAAM), April 2019

Revolutionary Approach for Smart Washing

Machine using Machine Learning & Artificial
Intelligence
Naveen Ananda Kumar J*, Srikaanth GV**, Kannika Nirai Vaani**
*
Institute of management,
Christ (Deemed to be University) Bangalore
**
Institute of management,
Christ (Deemed to be University) Bangalore

Abstract- Smart homes have started reshaping the place where people live and a smart washing machine
is an addition to it. The fast pace of our life has led to automation of many of the household items. The
next step in evolution after automation is smart and sustainable machines. In this paper we are introducing
a ground-breaking concept of smart washing machine which analyzes the clothes and based on their
specifications chooses the most appropriate type of wash for the clothes. This will reduce the fading of the
clothes and can reduce the usage of soap than the conventional washing machine. It uses Artificial
Intelligence and Machine Learning for enhancing the use of washing clothes more efficiently and
effectively. By using less water than the conventional washing machine, this technology is paving a way
to sustainability. Cameras, RFID patches, RFID reader are some of the technologies used for analyzing
the material of the cloth. Using AI and ML we can choose the suitable type of wash for the clothes. This
will increase the life span of clothes by mixing the right amount of detergent which in turn will reduce the
water used. This will lead to a smart and sustainable lifestyle.

Index Terms- Artificial Intelligence, Machine Learning, RFID, Smart washing machine, Sustainability

I. INTRODUCTION
rtifical Intelligence and Machine Learning are the new emerging trend in the home appliance
A industry. Homes are getting integrated with technology with almost everything in the house which is
taking us to the future. Consumers are becoming more tech-savvy and their expectation rises with the
advancement in technology [1].
The current washing machines have different washing method for textiles, which will not allow clothes to
get knotted, disarranged or entangled by intensive agitating cycles [2]. Damping the vibrations of the
drum of the washing machine is one of the latest designs implemented in the current washing machine
products [4]. [10] speaks about how only three wires are required for communicating inside the machine.
The network can be used for communicating via home network to manage peak energy usage and add a
new level of troubleshooting and field upgrades. A smart washing machine is a device which learns from
the previous data and improves the quality of the next wash [6]. Nowadays the conventional washing
machine has changed to a machine that can not only wash but also dry the clothes.
We are proposing a concept project where in, we are developing a smart washing machine that interacts
with you and is more sustainable. The usual problem with traditional washing machines is that different
materials of clothes need different types of wash. Using the same type of wash for different cloth
materials damage the clothes and its durability decreases. The amount of soap that is used to wash also

365
Alliance International Conference on Artificial Intelligence and Machine Learning (AICAAM), April 2019

plays an important role in durability of the clothes [7]. Conventional washing machines tend to use a lot
of water than required while washing the clothes.
The smart washing machine analyses these factors and accordingly washes the clothes without damaging
them and by using less water. It analyses the data of the clothes and accordingly chooses the appropriate
type of wash. This solves most of the above-mentioned problems faced in a conventional washing
machine.

II. CURRENT SCENARIO AND IT’S PROBLEMS

A. Present Scenario
In the recent times, textile industries have evolved to smart clothing. These involve smart clothes which
has various sensors in the fabric which gives various inputs such inbuilt music players, health monitoring
sensors, etc. [9]. Many of the textile manufacturing companies are adapting RFID technology by
embedding RFID tag in their products to track them from the warehouse to the customers [8]. The RFID
tag contains all the details about the product so it is easier to keep track of the products. [3] compares on
how much water is being used and how much water is actually required for washing.
[5] talks on how the data can be gathered from the washing machine. It also talks about bidirectional data
flow between washing machine and cloud for analytical purpose. A survey conducted by IBM found that
99.9% of the customers use only 3 of the many wash modes. This made the washing machine
manufacturers to cut down the uses of other wash modes which reduced their engineering and R&D costs
[5].
B. Research Gap
Washing machines have not evolved parallelly with other home appliances in the area of cloud
technology. Other home appliances such as television has moved from CRT TV’s to LED to Smart TV’s,
Refrigerators have moved to smart technology fridges and many more. There are lot of research papers on
the technical aspects such as reducing the sound and damping of the washing machine. Most of the
research papers are based on the automatic washing machines. This paper aims to bring washing machine
into the smart home sector.
C. Problem statement
To develop a smart washing machine, which analyzes the data about the clothes, then selects the most
appropriate method and uses minimum amount of water for the wash. It also uses the appropriate amount
of soap, so that the color of the cloth does not fade faster.
D. Objectives
This paper is intended to integrate AI and ML into washing machine so that it identifies the type of cloth
using RFID reader and based on that chooses the appropriate mode of washing. Based on the selected
mode, appropriate amount of soap and water is to be added with the right temperature.

III. PROPOSED MODEL

A. Components Used
A.1. RFID tag
It has all the details about the clothes such as the color, material of the cloth, the temperature at which it
should be washed and the appropriate type of wash.

366
Alliance International Conference on Artificial Intelligence and Machine Learning (AICAAM), April 2019

A.2. RFID Scanner

The RFID scanner scans the RFID tag which is embedded into all the clothes. Then it sends the
information to the microprocessor.
A.3. Microprocessor
The microprocessor collects all the data and sends it to the cloud. Based on the information it gets from
the cloud it starts with the washing process accordingly. It controls all the sensors in the smart washing
machine.
A.4. Sensors
A PH sensor is also fit to identify the PH level of the water entering the washing machine.
Water level sensor- To sense the amount of water to be filled.
Detergent level sensor- A sensor to measure the amount of soap present and the quantity that is dispersed
into the water for washing.

Fig.1.1 shows the components and flow of data from scanning clothes till washing of clothes in the smart
washing machine.

Smart Washing
Smart Washing Machine
Machine

Smart washing machine ID

Clothes with RFID reader ML and AI based algorithm

RFID tag in Cloud
Sensor

Weight gauge

Microprocessor

Fig.1.1: Architecture of Smart Washing Machine

process
B. Working process
The smart washing machine is first connected to the home WIFI network. Then the clothes are put in the
smart washing machine drum where the RFID tags embedded on the clothes are scanned by the RFID
scanner. This gives the details of the cloth such as the color, material type, the temperature at which it
should be washed, type of wash and weight, etc. The microprocessor takes the data of clothes from the
RFID scanner and send it to the cloud. The cloud which has machine learning and artificial intelligence
algorithms process the data and sends the appropriate temperature, the amount of detergent soap, the
amount of water required and best appropriate mode for the wash. This information is then sent to the
microprocessor, based on that it chooses the amount of detergent, water and the temperature of the water
and then adds it to the drum. Based on the type of wash, the microprocessor chooses starts the process of
heating the water and soaking the clothes for certain period. If there are 2 or more clothes which have to
be washed differently, a notification is sent to the mobile saying that there is a contradiction. So, the user

367
Alliance International Conference on Artificial Intelligence and Machine Learning (AICAAM), April 2019

can remove the cloth, or the user can ignore the message, which the washing machine automatically
chooses the best type of wash for it. After taking all these into considerations, it starts the wash. Once it
has finished washing, it will send a notification to the user that the wash is complete. The smart washing
machine then takes a feedback from the user and then sends the results to the cloud. These results will be
analyzed by the AI and ML which will use it for future reference. Fig.1.2 shows the working processes of
the washing machine.
Clothes (RFID
tag)

RFID scanner

Sensor (Detergent, Microprocessor Cloud

Water level)

Sensors

Selecting best type Choosing Selecting

appropriate appropriate quantity Selecting right
of wash amount of soap
temperature of water

Washing of
clothes

Notification that
washing is finished

Fig.1.2: Flowchart of working procedure

processprocess
B.1. Working of detergent soap sensing unit
The microprocessor checks the quantity of the detergent soap left in its storage. If it is less, it sends a
notification that the detergent level is less so that the user can refill it. After this the microprocessor
calculates the amount of soap and water required for the wash. Fig.1.3 shows the working of the detergent
soap storage in the smart washing machine.
Cloud

Electronic
Detergent sensors Microprocessor sensors valve
Storage

Information
about
clothes
368
Fig.1.3: Working of the detergent soap
sensing unit
Alliance International Conference on Artificial Intelligence and Machine Learning (AICAAM), April 2019

IV. EXAMPLE
A. Conventional Washing Machine
Let’s consider a situation where Ms. Padveeni is using an Automatic Washing Machine. She puts
different types of clothes (4 cotton, 1 wool and 1 silk) for wash in the washing machine. She sets the
temperature to 50 degree Celsius. She puts an approximate level of detergent soap in the machine. She set
the mode in Cotton mode because majority of the cloth put for wash were cotton clothes. Then started the
wash. After the wash when she removed the clothes, the cotton clothes were washed properly, but she was
not happy with the condition of the wool and silk clothes. This is because, wool cloth should have been
washed in a different temperature than cotton or silk and needs special care while washing. The silk cloth
requires very less amount of detergent soap than the other type of clothes.
B. Proposed model (Smart Washing Machine)
Let’s consider a situation where Ms. Padveeni is using the Smart Washing Machine. She puts different
types of clothes (4 cotton, 1 wool and 1 silk) for wash in the washing machine. The RFID scanner reads
the RFID tag of all the clothes put in the washing machine. It sends a notification that there are silk and
wool clothes which cannot be washed with the same setting as cotton clothes. She ignores the warning
and presses the start button. The Smart Washing Machine then finds the best wash for this set of clothes
i.e. it sets the temperature between 35- 40 degree Celsius.
It uses the approximate amount of detergent soap and correct level of water required for the wash. This
pattern of clothes and the appropriate type of wash gets stored in the cloud, so that the cloud can use the
same data for the next wash for the same type of pattern. After the wash, the Smart Washing Machine
collects the satisfaction level of the user and if there is any problem with clothes. This data gets stored in
the cloud. So, when the same pattern comes for wash next time, ML and AI can analyze the satisfaction
level and the type of wash used and corelate it. So, this can make the next wash more efficient and
productive.

V. FUTURE WORK
This proposed model can be used a base model for finding out the actual water requirement used by the
washing machine and also the amount of detergent soap which will be used. This can help the user
quantify the amount of detergent soap to be purchased. This can be used by marketers for finding out the
amount of detergent soaps required by every households and based on the consumption pattern they can
maintain the stock accordingly.

VI. CONCLUSION
This novel concept which is an integration of AI and cloud-based computing has a bright future and is
bound to branch out into various other dimensions of the home appliances especially with smart homes
which are becoming more common. The smart washing machine will enable its users for a smarter and
more sustainable use of water and detergent soap. As this washing machine programs and reprograms
itself based on Machine Learning and Artificial Intelligence algorithms in the cloud, preprogramming of
the washing machine is not required. This reduces the manufacturing and R&D costs for the
manufacturers. This Smart washing machine will open new opportunities and take smart homes
appliances to a new revolutionary level.

369
Alliance International Conference on Artificial Intelligence and Machine Learning (AICAAM), April 2019

ACKNOWLEDGMENT
We thank our colleagues from Institute of management, Christ (Deemed to be) University and Alliance
University who provided us with insights which was a great assistance in shaping our research. We would
like to express our gratitude to Prof. Dr. Lakshmi Shankar Iyer, Head of Specialization - Business
Analytics, Institute of Management and Dr. Manohar Kapse, Assistant Professor, Faculty- Business
Analytics Institute of Management for sharing their pearls of wisdom and for reviewing our research and
giving us key inputs.

REFERENCES
[1] Hultink, E. J., & Rijsdijk, S. A. (2009). How Today’s Consumers Perceive Tomorrow’s Smart
Products. Journal of Product Innovation Management, 26(1), 24–42. https://doi.org/10.1111/j.1540-
5885.2009.00332.x
[2] Islas, F. (2017). | mit kudo tori tutumie trolloni matematik, 1.
[3] Kiruthika, J., & Arulanantham, D. (2016). Making Washing Machines Smart through IoT, (April), 1–
4.
[4] Kolhar, S. S., & Patel, D. R. (2013). Optimization of a Drum Type Washing Machine By Analytical
and Computational Assessment, 4(6), 2759–2763.
[5] Locke, Dave and Lewis, K. (n.d.). Connected homes use case: Using washing machines to explain the
IoT. Retrieved March 13, 2019, from https://www.ibm.com/blogs/internet-of-things/washing-iot-
solution/
[6] Milasi, R. M., Jamali, M. R., & Lucas, C. (2007). Intelligent Washing Machine : A Bioinspired and
Multi-objective Approach, 436–443.
[7] Nakajima, R., Asano, Y., Hida, T., Matsumoto, T., Nakajima, R., Asano, Y., … Study, A. (2016). A
Study on the Effect of Dirt on an Inspection Surface on Defect Detection in Visual Inspection
Utilizing Peripheral Vision To cite this version : HAL Id : hal-01417598.
[8] Nayak, R., Singh, A., Padhye, R., & Wang, L. (2015). RFID in textile and clothing manufacturing :
technology and challenges. Fashion and Textiles. https://doi.org/10.1186/s40691-015-0034-9
[9] Suh, M., Carroll, K., & Cassill, N. (2010). Critical Review on Smart Clothing Product Development,
6(4), 1–18.
[10] Vanzoest, D., & Otten, D. (n.d.). Implementation of a Simulated Washing Machine with an Integrated
Control Panel and Intra-Appliance Local Interconnect Network Bus.

370