Instrument Panel Testing using Automated Panel key controller

CASE STUDY SNAPSHOT

Customer : A multinational in the domain of fire and safety catering to multiple geographies
Size : > 10000
Project vertical : Enterprise Security
Challenge : Need an automated Instrument Panel Testing system that can also be used for regression testing automation
Solution : Develop a TestBot powered Automated Panel key controller
Services Availed :  TestBot, AI Inference Engine, Electro-Mechanical design services, Embedded Systems Testing Services
Tools and Technologies :
  • Framework: TestBot, AI Inference
  • MPU: STMicro STM32F3 MCU and Nvidia Jetson Nano
  • Programming Language: C, Java and Python
  • Front End: Eclipse

INTRODUCTION

This project was executed for an existing customer with a requirement of testing instrument panels with multiple test cases already framed for the qualification of the device in accordance with the industrial standards. Having a long working relationship with them, Embien understood their needs and developed this instrument panel testing solution.

CHALLENGE

Though the instrument panel had the manual testing procedure done earlier, the need for more accurate and routine regression testing during developments has raised a need for the automated testing. Also, the manual instrument panel testing had a less success rate for obvious reasons. The major challenge was to do non-invasive testing without modifying any aspects of the existing product that has multiple mechanical buttons and LCD displays.

SOLUTION

Further to the requirement understanding, Embien designed and developed a Panel Key controller and integrated with our TestBot framework and included various test methodologies in adherence to the use cases and compliance requirements.

Instrument Panel

Instrument panels, being typically simple to install and use, play an inevitable role in industrial automation and building automation. Generally, they have user facing interfaces such as keypads, LEDs, multi-line LCD displays, etc. On the machine side, it could be anything based on the target application – such as RS232, Ethernet, Wi-Fi, Analog/Digital IO’s etc. Some example instrument panels include those used in production management systems, life safety systems, industrial HMI’s, advanced PLCs, HVAC, and security access devices.

Based on the underlying application and the need to create efficient and safer systems, the instrument panels could be complex and powered by advanced hardware, software, and analytics. Embien technologies with more than a decade of experience in designing instrument gateways, BSPs/drivers and protocol stacks for instrument panels has been serving OEM’s manufacturing instrument panels.

Instrument Panel Testing

The Instrument panel under test (DUT) consisted of the following subsystems,

  • Master controller
  • Display card with multiple navigation buttons
  • Sensor interface card
  • An Industrial IoT Gateway

Master controller was the heart of the system which functions 24 X 7 X 365. It communicated with the sensor interface card, acquired the data, and ran complex analytics on the acquired data for faster outputs. Display card included either character or graphical display and supported getting inputs from user such as configuration inputs, acknowledgment and indicated alarm/notifications to the user via LEDs and warning messages. In addition to the local notifications, the panel also updated/logged the events in the customer cloud via a gateway.

Regression testing was a pain point for the customer, since all the functionalities and use cases needed to be tested repeatedly for all the changes in the firmware. One other non-functional feature included the ability to perform Over the Air (OTA) firmware update, whereby devices in the field across geographies had to be upgraded based on need basis. With such a large system, when an upgrade was to be initiated, there were multiple test cases that needed to be validated before the final build reaches the customer site.

Embien Technologies clearly understood the customer needs, their compliance requirements and designed an automated test jig specific to their use case and integrated it with TestBot. The custom Instrument Panel Testing solution had the following features.

  • Automated Panel key controller for key/button actuation
  • Camera-based machine vision system for inspection

These are captured in the below diagram and explained in detail in the following sections.

Panel Key Controller

Panel Key Controller


Automated Panel Key Controller

With many success stories in mechanical designs, Embien’s Electro-mechanical design team developed the complete system for the panel key control. Following were the major design key features of the Automated Panel key controller.

  • Solution / Concept

    3-axis coordinated motion which automatically operated the panel with the help of user-friendly software.

  • Design / CAD

    Selection of driving machines

    • Stepper and Servo motors were the most preferred options for such a precise positioning operation. As the stepper could be driven in the open loop condition and does not require any additional feedback for positioning other than homing sensor, stepper motors were selected by performing motor sizing calculations for each axis according to their boundary conditions.
    • Mechanical power transmission – The mechanical power transmission method was precisely selected with the various considerations like vibration, lifetime, maintenance etc. Ball screws were used as it is a most suitable option for such linear operations which can easily deliver the power with swift response.

    Guiding the motion

    • It was important to guide the motion and linear guide rail was used. The reason besides using linear guide shaft was that it can be used as a frame for the machine which significantly reduced the system weight.

    Panel operating mechanism

    • A mechanism was required to be operated at a slightly increased speed at least equivalent to the human operation. To achieve this, the direct driving method of rack and pinion wheel was selected.

  • Validating the design

    Validating and optimizing the design was one of the important steps in a product life cycle management. All designed parts were simulated and analyzed using the Ansys software and the parts are optimized accordingly.

The panel key controller was designed accordingly that it should support all the variants where instead of the character LCD, TFT LCD will be available with touch feature. The key actuation unit had been designed in such a way that it can be converted into a stylus for actuating the touch panel as well. The entire system was driven by a small STMicro STM32F3 MCU, and a control interface provided over USB. Considering different types of panels with different dimensions and keyboard layout, the controller was designed to support all of them.

Camera based Visual Inspection

Our AI Inference solution was used to develop the visual intelligence for the system. Powered by Nvidia Jetson Nano, the camera-based machine vision system was provided for reading the characters/information from LCD/Graphics display and LED indications in the display and will be monitored for each sensor actuation. The Image processing was done primarily by OpenCV along with some machine learning algorithms such as YOLO and LSTM for pattern detection. Since various types of instruments panels were there, the algorithm was made generic and largely independent of camera position as well.

TestBot based regression testing automation

Both the custom test jig and camera system interfaced with our TestBot – automated testing framework – via agents. Test cases for Instrument Panel Testing have been reviewed and formalized by the customer and each test case was carefully converted into Python test scripts with all the prerequisites and expected results. All the test cases that require a key actuation, either to set the configuration or to process the information, leveraged the Panel key controller via USB interface. With this Automated Panel key controller set up, the test cycle was completely automated without any human intervention at all. It was also successfully used for regression testing automation and long run testing where the test cases need to be repeated for more than a day. With Python scripting, it was quick and easy for the customer to add new test cases as well.

BENEFITS

Thus, with TestBot and a customized Panel Key Controller, Embien was able to offer a unique solution for the customer’s problem of want of automated testing solution for their instrument Panel. Some of the advantages our solution offered are

  • Quick development using our ready to use TestBot and AI Inference Engine
  • Overall testing time was reduced to 22%.
  • Test coverage was increased to 100% without any exceptions.
  • Customized Electro-mechanical Test Jig that serves many instrument-panels.
  • Integration with Python based test scripts.

CONCLUSION

In this case study, we explored our expertise in developing the automated testing device for such an instrument panel testing using our TestBot framework. Saving significant development time, we are happy to be part of contributing to faster development and greener world.

Embien has been serving customers across domains like automotive, medical, industrial, manufacturing and defense for testing solutions such as instrument cluster inspection, IoT Gateway functionality checking, EOL validation of embedded devices, camera module validation, video conferencing system validation etc. Apart of being a end-of-line testing tool, TestBot has often proved to be a valuable regression testing automation tool aiding developers validate their releases thoroughly without spending precious time before releasing to QA.

Impressed by our work on face recognition-based access control on Jetson Nano?
Get in touch with our embedded ML development team today!