Predictive Maintenance of Industrial Equipments

Technicians can often tell when the machine doesn’t sound right. What if the machine themselves could alert about their health status?

Similar projects worth following
Problem Statement:

For years, manufacturers have been practicing a time-based approach to the equipment maintenance. They used to take the age of machinery as the factor for planning the maintenance routine. The older the equipment the more frequent maintenance procedures need to be carried out. The ARC group study states, however, that worldwide, only 18% of equipment has failed due to its age, while 82% of failures occur randomly. It proves that a time-based approach is not cost-effective – a piece of equipment gets maintained irrespective of the actual need.

The monitoring of equipment has previously been a potentially time-consuming and costly task due to manual inspections of each trap. When equipments are large in number the probability of failed equipment going unnoticed for an extended period of time is quite high.

Change the Planet with PSoC® IoT Design submission questions:

1. What planet-changing IoT project do you want to build?  

The aim of the project is to build a predictive maintenance system for any general Industrial Equipment by analyzing the time series data of temperature , Vibration and Acoustic signals and predicting the failure before hand using ML models on the cloud servers. 

acquire factory

Prime candidates for Predictive Maintenance include machines that:-
• Run 24 hours a day
• Perform functions that are crucial to the production process 
• Have high failure consequences
• Are expensive to maintain
• Could pose a risk to personnel safety or the environment.


Examples include machines used in continuous processes, e.g. plastic production, where a service interruption could ruin an entire production run. Those with high failure consequences include turbine generators and equipment located in hazardous environments. Also consider machines that experience frequent failures because of a tough operating environment — common in mines, cement plants, and fertilizer plants.

Initial signs of wear mechanical typically create high-frequency noise with could be detected using accelero-meters and microphone which are placed over the motor, data is collected and processed on the microcontroller itself using Fast Fourier Transform (FFT). Direct sending all collected data from the sensor node would put strain on the existing network bandwidth and also will lead to more power consumption. Hence the only FFT data is then sent to gateway. Looking at the steady-state vibrations, at the time of installation, thresholds were set.  

The data is then used to build a ML model , train it so that later on it could predict anomalous behavior of the equipment.  Then, connect the anomaly to the appropriate work task to minimize production loss, quality issues, or other collateral damage. 

Our system resorts energy harvesting to power the sensor nodes, with neither cabling for power supply (unadvisable in hazardous areas) nor batteries (which require regular maintenance).  The thermal dissipation around the motor is  used to generate a voltage through a thermoelectric generator. The voltage generated when fed into  low voltage boost converter can give regulated voltage output to charge a  supercapacitor. This in turn powers up the whole system in regular interval of time. 


• Compressed Fluid Leaks

• Vacuum leaks

• Steam trap failures

• Bearing condition monitoring

• Electrical arcing/tracking

• Fan and motor unbalance

The overall energy savings and minimum waste due this is invaluable towards developing  sustainable environment on planet. 

Transition from Reactive using SmartSense

2. Which Cypress PSoC® 6 Dev Kit would you like to use for the project and why? (you can use multiple kits)

I would like to use PSoC® 6 WiFi-BT Pioneer Kit (CY8CKIT-062-WIFI-BT) for this project as this the TFT display shield board include in the kit has 6-axis motion sensor, and a digital microphone which essential for monitoring the machine conditions. The Industry-leading CapSense would facilitate the Technicians to interact with the device even in harshest factory environment example wearing gloves . The display on the screen could be used alert the the coworkers nearby.  The data from the sensor is used transmitted via Bluetooth 4.1 to the gateway node and also the prediction results are being received from cloud to be displayed on the screen. 

3. How will you use AWS IoT or other cloud services in your project? 

I would like use AWS IoT service for the below mentioned tasks:

  1. The collection of data to build and train a model.
  2. The deployment of models back to the factory sensor nodes.
  3. The evaluation of data to perform local inference.

First we collect data from the machines or  equipment that you want to...

Read more »

  • Getting Stared with PSoC 6 WiFi-BT Pioneer Kit (CY8CKIT-062-WiFi-BT)

    Pratyush-Mallick08/22/2020 at 09:46 0 comments

    Thanks for the Hackaday community team, Cypress and Amazon AWS to make this happen. 

    I had a tragic incident with the PSoc 6 kit.  The first kit that I received turned out to be faulty , there were overheating issues which would end up the board being power off. But people at cypress were kind enough to send me another one . 

    Kit Contents

    The CY8CKIT-062-WiFi-BT package has the following contents, as shown in figure 1.

    • PSoC 6 WiFi-BT Pioneer Board
    • CY8CKIT-028-TFT Display Shield
    • USB Type-A to Type-C cable
    • Four jumper wires (4 inches each)
    • Two proximity sensor wires (5 inches each)
    • Quick Start Guide

    The Cypress PSoC® 6 WiFi-BT Pioneer Kit (CY8CKIT-062-WIFI-BT) is placed on a nicely finished non-conductive foam to avoid shocks, inside a rigid and high quality flip top cardboard box with magnetic catch for secure packaging. The build quality of package is simply amazing and high end. As device is ESD sensitive, it is highly recommended to take proper ESD precautions to avoid any permanent damaged to the device or any other malfunction (figure 2). Overall the it delivers a great unboxing experience

    A special note on the CY8CKIT-028-TFT Board , as is one of the main attraction of kit and houses a lost of sensors ( some of them don't have huge software support , so you might explore them on your own such as PDM examples and DAC examples)

    CY8CKIT-028-TFT Board : 

    • 2.4-inch 240x320 TFT LCD module
    • 6-axis motion sensor
    • PDM microphone
    • 32-bit stereo codec capable of
      • microphone
      • headphone
      • speaker amplifier.
    • 3.5mm standard audio jack with a provision of connecting both AHJ and OMTP headphones.
    • Headset standard selection switch.
    • An ambient light sensor
    • An 1.8V LDO for the digital supply of the audio codec

    The best thing I like about this board is there are abundant test points through out the board to debug if you run into any hardware issues. 

    The Kit comes pre-installed WICED WiFi demo.  This demo is meant to test the functionality of onboard WiFi-BT combo module along with CapSense touch slider, ambient light sensor and TFT display and use buttons. 

    In this demo, upon power up the kit becomes a WiFi hotspot with a SSID of “WICED Config” and password shown on the display. It hosts a wifi setup webpage where you can easily configure it to connect with one of the local WiFi network in STA mode. After successful configuration and  it hosts another webpage that displays the voltage of the ambient light sensor on the CY8CKIT-028-TFT shield. It also displays the current duty cycle of a PWM controlling the red LED on the kit. During this demo also it is also printing data over a UART to a terminal window.

    The first thing you need to do to tinker around the board is to upgrade the KitProg v2. KitProg is Cypress’ low-level communication firmware for programming and debugging - not available as a separate product.  Different software IDEs required different version of KitProg firmware. PSoC Programmer, Cypress Programmer or Modus Toolbox can automatically upgrade the KitProg2 to KitProg3 but automatic downgrade back to KitProg2 is not possible. PSoC Creater, Wiced Studio, Zerynth required KitProg2 where as Modus Toolbox will only work with KitProg3. 


    There a lot of tutorials on this on the web , hence I'm gonna go through this.  I have attached a list of resources below to help you out. 

    The Big Question : How to choose the right IDE?

    This question been haunting even after finishing the project. There are variety of IDEs that the PSoC support each one is suitable for specific application. I guess most of the time you would be juggling between IDEs . The cross-support for each IDE is available , but still is in the initial phase and for my experience wasn't much of a help. I would give overview of the options available : 

    Modus Toolbox

    (Highly Recommended) it is easy to use programming environment for PSoC6 series mcu. If someone has prior working experience...

    Read more »

View project log

Enjoy this project?



Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates