Predictive Maintenance on Azure IoT - …

1 Predictive Maintenance on Azure IoT. IoT Big Data . In Kee Paek, Cloud Solution Architect IoT Predictive Maintenance . 50%1 $630B . (2025 . ). Predictive Maintenance , . (OEE) 30%3 . 10 .. In contrast to a traditional preventive Maintenance system, Predictive Maintenance solutions enables customers to strategically plan their Maintenance tasks and group them in a way that allows them to perform the required Maintenance more efficiently This leads to even more dramatic savings in terms of labor and Maintenance costs . Melissa Topp, Director of Global Marketing, ICONICS. 1 McKinsey, The Internet of Things: Mapping the Value Beyond the Hype, 2015. 2 McKinsey, The Internet of Things: Mapping the Value Beyond the Hype, 2015. 3 GE, The Impact of No Unplanned Downtime, 2014. 4 GE, The Impact of No Unplanned Downtime, 2014. Address business needs across a range of scenarios.

2 (uptime) .. ADD BEST. Maintenance PRACTICE. INITIATED. | . | . Gain visibility into product Utilize service and performance data to Offer Predictive and proactive services Identify trends and potential growth performance and enable workflows to proactively detect product failure. to address customer and product opportunities using customer respond to changing conditions. needs. sentiment and product usage data. Jabil .. Jabil wanted to better predict Jabil was able to Predicted machine processes errors or failures on the transform their that will slow down or fail with assembly floor before they manufacturing production an 80% accuracy occur, saving customers' time lines with advanced Reduced costs of scrap and re- and money. analytics solutions built work by 17%. on Microsoft technology. Delivered energy savings of 10%. We'll be able to improve our efficiencies, cut costs, and decrease our lead times, which tie directly to our customers' requirement to increase flexibility.

3 Matt Behringer CIO, Enterprise Operations and Quality Systems, Jabil . Source of failure can be introduced at multiple stages but cannot be detected until it is powered-up for testing at the end Inspection steps along the SMT line cannot always Product detect the quality issues quality not acceptable Jabil As-Is . Continuous requirement to increase yield, reduce amount of scrap and re-work Reduced manufacturing cycle time Traditional inspection techniques for ensuring quality quickly becoming Higher cost of wasted materials, time and resources outdated Inability to address customers' critical requirement for speed to market with more one-off production runs Adding more equipment and people to existing manufacturing processes would not have significant impact on increasing throughput . 1 1 0 1 0 1 0 1 0 1 0 0. 1 0 1 0 1 0 1 0 1. 0 1 0 1 0 1 0 1 0 1 Historical 01 data 0 0 0 1 0 0 0 1 0 0 0 1 0 0 0 1 00 0 1 0.

4 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0. 0 1 0 data 0 1 0 1 Supplier 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 10. 0 0 1 0 0 0 1 00 0 1 0 0 0 1 0 0 0 1 0 0 0 1 0 0 0 1 0 0 0 110. 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 1 0 0. Azure 01 0 1 01 0 101 0 1 01 0 1 01 0 1 01 0 101 0 101 01 01 0. 0 0 Services 1 0 0 0 1 00 0 1 0 0 0 1 0 0 0 1 0 0 0 1 00 0 1 0 0 0 1010 10 Our rate of rejection has 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 00 11 01 00 1 011 0 1 0 1 0 1 0 1 0 1 0 1 0 10 1 0 10 decreased dramatically now 1 0 10 1 10. 0 PREDICTED. 1 10010 0 100 010 100 0 0 10 0 0 10 0 1100 0 1 0 0 0 10 0 0 10 0 0 10 0. 0. 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 PREDICTED0 1 0 FAILURE 0. FAILURE that we can predict these Customer 0 0 0 1 0 0 0 1 0 0 0 1 0 0 0 1 0 0 0 1 0 0 0 1 0IN 0 NEXT 10. 0 48 11 01 0 110 01 11 1 0 1 0 1 0 1 0 1.

5 0 0 0 0. 01 0101 101 0101. Recommended Maintenance 1 data HOURS 0 0 0. in next 48 hours failures early in the process 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 10 10 10. 1 0 1 0 Production 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 10. 1. 0 0 1 0 data 0 0 1 0 0 0 1 0 0 0 1 0 0 0 1 0 0 11 01 0. 01. 00. 01. Variation: 11% (tolerance 11%). 10 Tuesday October PREVENTATIVE. Maintenance . 0 0 1 0 0 0 0 0 0. 0 1 1 0 0 10. 0 01 1 1 10 1 1 1 0 01 00 0. 10 0. 7:00 AM FOR TOMORROW. 01 0 0 1 10. 10. 00 1 0 1 1 0. 10 1. 10. 0 01 01 00. 01 Vibration frequency: Too high 10 0 10 0 0 01. 1 0. 0 1 0. 1. 0 1 0. 00 1 0 1. 0 1. Bit wear: High CRM ERP MES SPC Other systems Result in plant 2: Failed . Jabil wanted to better predict errors or failures on Jabil was able to transform their manufacturing Predicted machine processes that will slow down the assembly floor before they occur, saving production lines with advanced analytics solutions or fail with an 80% accuracy customers' time and money built on Microsoft technology Reduced costs of scrap and re-work by 17%.

6 Delivered energy savings of 10%. ThyssenKrupp Elevator .. ThyssenKrupp wanted to Microsoft technology Increased elevator uptime better compete in their enabled ThyssenKrupp to Reduced costs for ThyssenKrupp industry by offering monitor products via a and its customers dramatically increased real-time dashboard and uptime, taking preventative instruct technicians on Developed real-time data Maintenance a step further to optimal Maintenance visualization and awareness of Predictive and even activities through issues preemptive service. dynamic Predictive models. We wanted to go beyond the industry standard of preventative Maintenance , to offer Predictive and even preemptive Maintenance .. Andreas Schierenbeck CEO, ThyssenKrupp Elevator .. Thyssenkrupp wanted to better monitor their more They connected thousands of sensors embedded in Reduced costs for thyssenkrupp and its customers than million elevators worldwide.

7 Lack of insight their elevators to the cloud to monitor real-time Increased reliability through Predictive Maintenance led to downtime and unpredicted failures in some performance and proactively address issues with and rapid, remote diagnostic capabilities of the world's most famous buildings Microsoft technology Scheduled Maintenance Hydraulic/Passenger Elevator Replace brake shoes Check AC motor for health ! NEXT STOP. ! Maintenance in progress Microsoft Offers Two Approaches to IoT Solutions PaaS Azure IoT Suite SaaS. PaaS. SaaS Microsoft IoT Central Microsoft Offers Two Approaches to IoT Solutions Azure IoT Suite Microsoft IoT Central To accelerate development of a custom IoT solution To accelerate time to market for straightforward IoT solutions Primary usage that needs maximum flexibility that don't require deep service customization Access to underlying Access to the underlying Azure services to SaaS.

8 Fully managed solution, PaaS Services manage them, or replace them as needed. underlying services aren't exposed. High. The code for the microservices is open Medium. leverage built-in browser based user experience to cust Flexibility source to be modified. omize the solution model and aspects of the UI. Medium-High Low Skill level Java or .NET skills are required to customize the solution back end. Modeling skills are required to customize the solution. JavaScript are required to customize the visualization. No coding skills are required. Preconfigured solutions implement common Templates provide pre-built models. Get started experience IoT scenarios. Can be deployed in minutes. Can be deployed in minutes. Pricing You can fine-tune the services to control the cost. Simple, predictable pricing structure. Azure IoT Suite - Accelerate Time to Value! Get started in minutes Fine-tuned to specific assets and processes Modify existing rules and alerts Highly visual for your real-time operational data Add your devices and begin tailor to your needs Integrate with back-end systems Azure IoT Suite Solutions What you get with the preconfigured solution?

9 Devices Azure IoT Suite Predictive Maintenance , Remote Monitoring Azure IoT SDK. (OSS). Linux, RTOS, mBed, Web/ Power BI. Mobile App Windows, Android, iOS. Storage blobs DocumentDB Existing Business Process ERP/CRM. IoT Hub Stream Event Hub Web Jobs Logic Apps Analytics Azure Machine Learning Active Directory Azure IoT Suite Reference Architecture App Service Optional Web App Tech Options AAD. User Management OAuth2 Providers Config. &. Rules Definition IoT Edge(s). IoTHub Spark Event Grid Data Xformation Strea Analytics Stream Analytics Notifications / Logic Apps Cloud Gateway ASA. Microservices (Rules/Insights). (Rules/Insights) Actions AF. IoT Device(s). Azure Storage Kubernetes Azure ML. TSI. Cosmos DB. Orchestration Service Fabric Cold Path Machine Warm Data Store Learning Predictive Maintenance project .. 1 2 3 4 5. Identify the Inventory data Capture & Model, test and Integrate into target outcome sources combine data integrate operations Azure IoT Suite solutions come with pre-built sample scenarios that include: Background information on the business need and objectives Simulated devices and sample data Pre-set rules and alerts, pre-defined dashboards and more Approach: Checklist for Machine Learning The better the raw materials, the better the product.

10 Data Question is measures Data is Data is A lot of sharp. what you accurate. connected. data. care about. Will be Predict Identifiers Failures Machine difficult to whether at the level are really information predict failure component X they are failures, human linkable to accurately with will fail in the predicting labels on root usage few examples next Y days causes information Approach: Data Science Process Analysis Dataset Publish Define Objective Data Sources Explore Data Machine Learning Approach: Data Sources for Predictive Modeling FAILURE HISTORY REPAIR HISTORY MACHINE CONDITIONS. The repair history of a machine, The operating characteristics of a The failure history of a machine or previous Maintenance records, machine, data collected from component within the machine. components replaced, Maintenance sensors. activities performed. MACHINE FEATURES OPERATING CONDITIONS OPERATOR ATTRIBUTES.