This post is authored by Danielle Dean, Senior Data Scientist Lead at Microsoft.

The aerospace industry involves many complex challenges including intricate logistics and lends itself to several very impactful applications that involve connecting assets and devices through the Internet of Things and deriving actionable insights from deep analytics.

The combination of IoT and advanced analytics allows businesses to improve their operational efficiencies, safety, and even open new revenue streams. Take, for example, modern aircraft engines, which are equipped with highly sophisticated sensors to track the health of these machines. By combining the data from these sensors with the latest analytics techniques, it’s possible to both monitor the aircraft in real time, as well as predict the remaining useful life of an engine component so that its maintenance can be scheduled in a way to prevent mechanical failures. To address this type of business need, earlier this week we released a public preview version of the Cortana Analytics Predictive Maintenance for Aerospace Solution Template. This template can be used to monitor aircraft and predict the remaining useful life of aircraft engine components. The template utilizes realistic publicly available data, and we were able to create this solution thanks to this repository and the donators of this data [1].

Predictive maintenance offers aerospace businesses a real competitive edge – a world in which you could not only predict equipment failures before they happen, but also systematically address them. When a business harnesses the power of data and advanced analysis, it can identify issues in advance, predict when equipment needs attention, and preemptively service that equipment.

Putting Data to Work Towards Predictive Maintenance with Cortana Analytics

Traditionally, companies have maintained business assets through manual processes that are time and resource intensive. But today, with the power of data ingestion, storage, processing and advanced analytics in the Cortana Analytics Suite (CAS), aircraft maintenance and other predictive maintenance scenarios are opening up to a world of new possibilities.

The Cortana Analytics Solution Template is complementary to the pre-configured solutions available today through the Azure IoT Suite, including remote monitoring and predictive maintenance. For businesses, this also offers an opportunity to drive commands to devices and integrate into existing business systems.

What’s Under the Hood

The Predictive Maintenance Solution Template in CAS brings together several Azure services. Stream Analytics provides real-time insights on engine health and stores that data in long-term storage for more complex, compute-intensive batch analytics. HDInsight transforms the sensor data at scale which is then consumed by Machine Learning to predict the remaining useful life of aircraft engines and components after each flight. Data Factory handles orchestration, scheduling, and monitoring of the batch processing pipeline. The solution template comes with a data generator, and by running it after using the 1-click deployment of the underlying data processing and analytics services, gives you an idea of how these different components can be used together to predict the remaining useful life of aircraft engines. Finally, a Power BI dashboard can be built on top of the pipeline to allow aircraft technicians to monitor the sensor data from an airplane or across the fleet in real time, using visualizations to schedule maintenance on engine parts.

Solution Template Architecture

Wondering where to get started with your own predictive maintenance problem? If you haven’t yet connected devices or assets directly to the cloud or to each other, you can get started within minutes with the Azure IoT Suite. If you’re looking for data processing and advanced analytics on top of your data, then the place to start is the Cortana Analytics Solution Template. Wherever your business may be in its journey, Microsoft’s leading IoT and advanced analytics offerings provide a simple and unified approach, with preconfigured solutions and templates that are designed to help you benefit quickly from new opportunities and transform your businesses through the power of actionable data-driven insights.

Danielle

References:

[1] “A. Saxena and K. Goebel (2008). “PHM08 Challenge Data Set”, NASA Ames Prognostics Data Repository, NASA Ames Research Center, Moffett Field, CA.”