IOT based fault monitoring device for Comsats water pumps

As the world's technology advances, the use of water pump in daily life and many parts of life, such as industrial operations, construction sites, and so on. Faults in water pumps occur and we are on detect these faults when any major damage occurs, which results in unscheduled downtime, increased m

Project Title

IOT based fault monitoring device for Comsats water pumps

Project Area of Specialization

Electrical/Electronic Engineering

Project Summary

As the world's technology advances, the use of water pump in daily life and many parts of life, such as industrial operations, construction sites, and so on. Faults in water pumps occur and we are on detect these faults when any major damage occurs, which results in unscheduled downtime, increased maintenance charges, breakdown of equipment that results in changing the equipment and most importantly so, in order to improve practicality, things should first be monitored; otherwise, a minor incident might result in serious destruction, which is not cost-effective. This initiative allows us to spot problems early in the process. So that the time and life span of the appliance can be enhanced.

Variations in current and voltage on the supply side can cause faults in water pumps. If faults are not detected on timely basis, they can cause significant damage to the water pump, leading to several other significant issues i.e.

• Wastage of time and money on pump maintenance
• Could result in permanent pump damage
• Could be dangerous for the operator

With current technological advancements electric appliances are increasingly moving in to the world of the Internet of Things (IoT). IoT based automation of the electric appliances significantly reduces operating expenditures when automation devices, sensors etc become Internet-enabled devices and the system working as one entity produces alarm and notifies user whenever the system detects abnormality in the behavior of the appliance. Monitoring health of electric appliances detects possible faults before any disaster happens. This project is an IoT based fault monitoring system that monitors the behavior of electrical magnitudes in water pumps in real time, evaluate acquired data, detect potential faults, and notify the user of the situation.

The fault detection system first automates the pump by the help of current sensor and environmental temperature sensor. The next phase is the prediction phase, fault is predicted at early stages and the system notifies the user to take precautionary measures.

Project Objectives

Following are the primary objectives of the project:

• To interface current sensor with water pump and analyze data acquired through it

           Non-invasive current sensor is interfaced with water                 pump and data is continuously acquired through the               sensor at different time intervals 

• To monitor the health of water pumps by the implementation of Machine Learning (ML) Algorithms

           Data acquired through current sensor is classified                   according to the max amperage that water pump can               bear as per water pump's name plate specifications.

            Then an algorithm is implemented on that data being              the training data and a model is made. Model is then              tested using the data other than training data

• To develop an automated premature fault monitoring and detection device

          Model is deployed and a premature fault is                              detected whenever there is an abnormality in the water            pump's current. Water pump's current is continuously              monitored and any abnormality in current is                              detected before any catastrophic damage or any                      damage which may breakdown the water pump or                  drastically increases its maintenance charges as well              as the time it takes to repair.

Project Implementation Method

     

System Architecture

Data Collection Unit

Non-invasive current sensor is interfaced with the water pump and data is collected at regular intervals. SCT-013-030 current sensor is used for this purpose. Data is collected and classified at this phase according to water pump's max amperage bearing limit. Current's allowable deviation is 5-10% of water pump's max amperage. Data is classified according to this limit.

 Model Training, Testing and Evaluation in Weka

Weka is a Machine learning tool is that used at this stage for evaluating the classification model having the highest accuracy. Collected data was fed into the software and different supervised machine learning algorithms were applied to the dataset. Different evaluation tests such as split test, 10 fold cross validation and repeated holdout method were done repeatedly individually on dataset when applying different algorithms such as J48, Naïve Bayes, K-Nearest Neighbor (KNN) and Logistic Regression (LR).=

KNN gave the greatest accuracy among all supervised machine learning classification algorithms.

Model Deployment

Deployment is the method by which you integrate a machine learning model into an existing production environment to make practical business decisions based on data. It is one of the last stages in the machine learning life cycle and can be one of the most cumbersome. This model is deployed and as a result human intervention free water pump fault monitoring and detection device is made. 

Block Diagram of the Proposed System

Components

Following hardware tools are used in the proposed project

• Raspberry Pi 4
• Biasing Resistors and Capacitor
• Burden Resistor
• YHDC Current Transformer SCT-013-030

 

 

 

 

 

Benefits of the Project

Advancement in technology has led to ensuring a better experience as this project does. It provides for the detection of any incident, as well as the collection and transmission of data to the cloud. Furthermore, the user is informed of the severity of the fault that will occur. Machine-Learning intelligently analyses the process, resulting in reduced time consumption and accurate and early fault detection. It uses less energy and runs more effectively. For domestic purposes, it is a very safe and reliable technology that does not require human intervention. The sytem can be used and modified for the fault detection of  many of the home appliances like refrigerators, microwave ovens etc.

The system also complies with the United Nations' sustainable energy goals.

Enhancing research and supporting domestic technological development is one of Objective of UN SDG 9. The sytem helps preventing domestic appliances from failure.

Targets of SDG 12 is to Support developing nations in strengthening their scientific and technological capability; develop and implement technologies to assess sustainable development impacts.

Technical Details of Final Deliverable

Block Diagram

For data collection SCT-013-030, which is a 30A non-invasive Hall Effect sensor is interfaced with water pump and data is collected. Collected data is used for training and a classifier is then implemented and prediction is done.

Final product will be a human intervention free device which will monitor water pump’s current on regular intervals and whenever there will be any abnormality in pump’s current i.e. deviation of current from its rated value is more than 5-10%, user is notified that pump needs maintenance or repair your pump before any major damage occurs to pump.

Collected Data Sample

SCT-013-030 current sensor was interfaced with a 2 hp centrifugal water pu,png a max amperage  of 11A. 

Above collected data shows that the water pump is already running at a current much greater than its rated value. This is where the problem lies. Water pump should go through maintenance phase at this stage to prevent any downtime, increased maintenance costs and major damage occuring to equipment or causing any harm to humans. Our system will detect this abnormality and user will be notified about the condition of the pump timely.

Final Deliverable of the Project

Hardware System

Core Industry

Energy

Other Industries

IT

Core Technology

Artificial Intelligence(AI)

Other Technologies

Internet of Things (IoT)

Sustainable Development Goals

Industry, Innovation and Infrastructure, Responsible Consumption and Production

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com