Predictive Maintenance IoT Device Development using WPF & Qualcomm

In this case study, we demonstrate the development of a Predictive Maintenance IoT Device powered by Qualcomm hardware, integrated with a WPF desktop dashboard , Flutter mobile app, and AWS IoT for real-time data processing and connectivity.

The issue of industrial downtime presents many difficulties for those who work within the manufacturing industry. Manufacturing firms incur production losses, higher maintenance costs and face risks regarding employee safety as a result of unexpected equipment breakdowns or failures.

To solve these problems, we created a Predictive Maintenance IoT solution that will be based upon Qualcomm hardware, complete with monitoring dashboards developed with Microsoft's WPF development platform and a mobile application created using the Flutter programming language. The IoT Solution will provide real-time insights and alerts.

A global manufacturer is experiencing frequent breakdowns and operational inefficiencies with their manufacturing equipment because of equipment failure.

It was essential for them to have a predictive maintenance solution to provide real-time equipment monitoring, identify any irregularities, and produce useful information to enhance performance and reduce costs.

To assist the company in developing the overall integrated smart industrial device and delivering to its customers an overall solution, we developed the hardware, firmware, software, cloud, and manufacturing of the proposed system.

The overall goal was to create and implement an Industrial Predictive Maintenance (IPM) solution, using Qualcomm processors and focused on achieving the following objectives:

• Design Industrial-grade Hardware and Custom PCB
• Firmware and Embedded Systems Development to capture real time data
• Build Rugged and Durable Industrial Housing for Harsh Environments
• Develop Cloud Data Storage, Analytics and Device Management Infrastructure
• Cross-Platform Software (Web, Desktop, Mobile)
• REST API and SDK for Integration with SCADA, ERP, and IoT Solutions
• Over-the-Air (OTA) Firmware Updates and Remote Device Management

The complete integrated solution provides a single end-to-end solution where all components such as sensors and cloud analytics can work together to provide optimal performance for Industrial Operations.

Product Type: I.S.M.-E.P.M. (Intelligent Systems for Machinery Equipment Predictive Maintenance)

Primary Purpose:

To monitor machinery on an industrial site, detect deviations, predict upcoming failures, and provide actionable information that will minimize downtime and improve operational efficiency for that machinery.

Key Features:

Real-Time Equipment Monitoring:

Measurement of vibration, temperature, power consumption and operational parameters on a timely basis.

Predictive Analytics:

AI/ML to determine the early stages of failure on machinery.

Cross-Platform Applications:

Mobile, web, and desktop access for monitoring, reporting and notification.

Cloud Capable:

Data is securely stored, can provide historical trending, capable of managing devices.

API and SDK:

System integrator capability to integrate SCADA systems, ERP systems and third-party analysis packages.

Remote Update of Firmware:

Allows for over-the-air (OTA) updates to enhance performance of product.

Industrial Standards:

Functionally IP65 rated, delivered with heat resistant materials and compliant to EMI and/or EMC regulations.

Processor Used: Qualcomm Snapdragon 410E Industrial Platform

Reason for Selection:

Excellent Performance:

Higher Performance due to quad core ARM Cortex-A53 CPU to efficiently process multiple Sensors Streams, AI Inference, & Real-Time Analytics.

Connectivity Options:

Onboard WiFi/BLE Connect to LTE/5G optional modules ensures seamless connectivity in Industrial Environments

AI/ML Capabilities:

On Device AI Processing for Predictive Maintenance Algorithms without excessive latencies and Dependence on Cloud.

Highly Secure:

Secure boot, Hardware Encryption, and Trusted Execution Environment Supports the protection of Sensitive Industrial Data.

Scalable/Dependable:

Meets the stringent requirements for Long Lifecycle and Extended Operating Temperature Range for Industrial Environment.

Eco System Support:

Predictably Assured, and Fully Supported SDK's and Development Tools Available to Help Engineer Faster Integration of Firmware, Software, and Developing AI Algorithms.

1. Hardware & Circuit Board Layout

Custom Build Multi-layer Board for Analog and Digital Signal Separation; Reduces Noise and Increases Accuracy of Sensor Data.

Integrated Sensors:

Vibration Sensors, Temp Sensors, Current Sensors, Volt Sensors, Accelerometer (for Health Monitoring of Equipment).

Connectivity Methods:

Wi-Fi, BLE, Ethernet, Optional LTE/5G for Remote Monitoring

Power Supply:

Industrial Grade with Surge Protection and Battery Backup For Continuous Power

Enclosure:

Rugged IP65 Design (for Dust, Moisture, and Impact Resistance)

Compliance:

EMI/EMC Certification & Industrial Standards Compliance

2. Embedded & Firmware Development

Firmware utilizing RTOS:

Allows for the use of multiple tasks to perform real-time sensor monitoring and edge processing using AI algorithms

Sensor Fusion Algorithm:

Uses more than one sensor to provide accurate detection of any equipment behaviour and errors in real time

All communication protocols supported:

MQTT over TLS, HTTP REST PROTOCOL, and Modbus TCP for SCADA integration

Over-the-air (OTA) Software updates:

Updates firmware for devices that were deployed remotely

Power optimization strategies:

Utilizes low power states during periods of inactivity which reduce energy consumption

Error Handling and Recovery:

This feature allows the device to maintain reliability while operating under extreme industrial conditions.

3. Industry Design & Housing

• Durable Industry Housing - Metal/Plastic resistant to temperature extremes for industrial environment.
• Modular Design - Allows easy change out of sensors and ease of service and upgrade.
• Visual Indicators - LED lights indicating operational status and warning.
• Easy to Install - Designed to be installed quickly into factory machines.

4. Cloud and Backend Architecture

• AWS IoT Core/Azure IoT Hub provides device management and telemetry services to manage devices on a large scale.
• Hybrid (relational and time-series) database systems are being utilized to store both historical and real-time data, as well as enable analytics.
• An analytics engine that uses predictive AI algorithms to identify abnormal patterns in both edge processing and cloud-based computing.
• The ability to send real-time alerts and notifications via Desktop, email, mobile and web applications.
• The use of web-based dashboards to provide management and operational support to achieve same.

5. Software Development

• Mobile Applications: iOS and Android apps provide real-time monitoring, alerts, and analytics using Flutter
• Desktop Applications: Provides a centralized control system with the ability to operate in offline mode
• Web Dashboard: Device registration, performance metrics, predictive analytics, and historical reporting
• Desktop Application: Advanced analytics, logging, and configuration management
• SDK & REST APIs: Enable third-party integration with SCADA systems, ERP platforms, and industrial IoT ecosystems
• Security: Role-based access control, encrypted data transfer, and secure authentication

PCB Assembly:

Produce SMT & Thru-Hole Components with an industrial focus that will stand up under rigorous testing and performance.

Functional Testing:

Confirm sensor accuracy, sensor-to-sensor connectivity, and predictive analytics performance.

Environmental Testing:

Heat, Vibration, & Humidity testing to simulate a manufacturing environment.

Batch Quality Assurance:

Automatically programmed firmware flashing to provide each unit with calibration and quality assurance verified.

• Delivered a fully operational predictive maintenance device powered by Qualcomm technology
• Provided real-time monitoring and predictive analytics with excellent accuracy
• Enabled remote monitoring and management via cross-platform applications
• Established scalable cloud infrastructure capable of supporting hundreds of devices distributed across multiple plants
• Reduced unplanned downtime by 20-25% through predictive maintenance algorithms
• Developed SDKs and APIs for easy integration with SCADA, ERP, and third-party industrial systems

• There needs to be a collaborative development process for the full production of Industrial IoT, which includes: Hardware, Firmware, and Software and Cloud Teams.
• Qualcomm processors allow for high performance capabilities for processing solutions to enable AI; and, provide robust connectivity which are all necessary elements in completing production of Industrial IoT Devices.
• Using Industrial grade Casings, PCB, and Sensor Integration are all integral to ensuring that products will provide a high level of sustainability and compliance when producing under adverse conditions.
• Using Cloud infrastructure and Cross-Platform Applications provide the capability to scale on demand, provide predictive analytics and real time monitoring among other examples
• The use of OTA Updates, SDK's, and API's allows for long-term device life cycle management and the ability to integrate devices into larger Industrial Ecosystems.