How Storm Reply Enables Industrial IoT and Predictive Maintenance at Schenck Process Group with AWS IoT
By Lennart Neumann, Sr. Consultant – Storm Reply
Many of the long-established industrial enterprises begin digitizing their manufacturing devices to improve visibility into their processes and the effectiveness of their production lines.
This endeavor requires immensely stable and future-proof IT solutions suitable for the complex logic and business structure of an organization that has often been developed over decades.
At Storm Reply, an AWS Premier Tier Consulting Partner and Managed Service Provider (MSP), we find that when it comes to the Industrial Internet of Things (IIoT) and Industry 4.0 platforms, our customers often face the hardest decision right from the start: build vs. buy.
This decision is tough because both options can have real and critical drawbacks for most organizations.
Choosing to build a platform enables almost infinite possibilities in terms of functionality, but leaves customers with a huge effort in development and ongoing maintenance of the platform.
On the other hand, going for a ready-made software-as-a-service (SaaS) solution for Internet of Things (IoT) enables the stability, scalability, and low maintenance customers are looking for. Unfortunately, while all solutions in this category offer some customization, they are often so limited that the business logic and data structure the organization has been working with for a long time can’t be built into them.
In this post, I will discuss how Storm Reply used Amazon Web Services (AWS) IoT services to implement a solution that is highly scalable, easy to maintain, and optimized to satisfy custom requirements for Schenck Process Group, a global market leader in measurement and process technology.
Schenck Process—a Storm Reply customer offering B2B customers predictive machine maintenance and advanced industrial monitoring— needed a scalable IoT platform capable of managing thousands of devices. They also needed an application to integrate the data structure they had been working on for years.
At Storm Reply, we believe customers should not have to compromise on either. That’s why we designed a solution to collect, store, and view sensor data alongside their context within an organization (plant, production line, machine part, and so on) to create real business value for the customer.
Our approach uses AWS IoT and AWS IoT Greengrass as a base to build a stable and scalable solution for the heavy lifting that comes with managing devices and connections at scale and preprocessing data at the edge.
Storm Reply’s overall solution relies on AWS IoT for the groundwork of an IoT solution: data ingestion, device management, device monitoring, and secure connections to the cloud. This leaves us to build high-quality software on AWS that meets client demands, enabling them to innovate and generate value.
Figure 1 – Schenck Process IIoT platform reference architecture.
Edge Device Management and Data Ingestion
The Schenck Process IIoT platform requires measuring a vast range of data points from many different sensors to offer predictive and data-driven maintenance to their clients. These sensors are positioned on machines across the globe, often in remote locations.
The data ingestion processes and device management are therefore required to be immensely scalable and reliable. Furthermore, a broad and ever-growing portfolio of machines with different connection technologies and preprocessing needs must be supported.
Essentially, our approach is to leverage AWS IoT, AWS IoT Greengrass, and their native MQTT connection handling to make the management and connection of edge devices as easy and low-maintenance as possible.
This has multiple advantages over building MQTT implementations on your own:
- The management platform to monitor your connected devices doesn’t have to be set up or built. It’s already there in the AWS Management Console.
- AWS IoT Greengrass on the edge device allows you to control the ingestion of data as you see fit. For example, we choose to send regular, time-based sensor data over MQTT and process them with IoT rules and AWS Lambda functions, while pushing files of greater sizes to Amazon Simple Storage Service (Amazon S3).
- The setup of new machines can be automated almost entirely as long as a predefined certificate is present on the machines.
Dedicated Storage Solutions for Ingested Data
Due to Storm Reply’s experience with various customers, we have learned there are multiple types of data needed to be ingested into the Industrial IoT platforms differing in frequency and size.
We differentiate between time series data pushed regularly (often near real-time) and bigger chunks of raw data being transmitted less frequently. Storing and maintaining this data and their access to it calls for respective solutions built to satisfy the different requirements.
For time series data, we use Amazon Timestream, which helps us keep the scalability of our platform because it’s a serverless service. We recommend Timestream over a traditional relational database because of the performance that’s required by the time-based queries needed for line charts and trend diagrams.
For flat, non-time series data, the choice is Amazon DynamoDB, as its speed and serverless nature are a great fit for our platform requirements. Furthermore, a NoSQL database enables us to adapt our frontend application more easily, as requirements change often while our clients are learning about their data and discovering new functionalities.
Big data and business intelligence usually play an essential role in an Industrial IoT application. We store all raw data on Amazon S3 by default because the costs are negligible, and it enables customers to query the data in the future using Amazon Athena.
Custom Web Application for Data Visualization
At Storm Reply, we have seen IIoT customers have specific requirements for the way data is visualized and used by their own organization and their B2B customers. This is why we recommend building the web and mobile applications in close alignment with clients to cover all requirements, learn quickly from the data, and innovate without restrictions.
To integrate and build useful APIs for the different data sources, we need a powerful tool for data integration—like GraphQL with AWS AppSync. This approach can unify multiple databases and technologies in one API, and gives us speed and flexibility when it comes to showing complex data in web applications.
You may find that your organization has to make great sacrifices for an all-encompassing Industrial IoT platform.
Due to its serverless nature, AWS IoT is immensely scalable and stable. This, paired with the extensibility that comes with the AWS Cloud, means organizations don’t have to compromise on stability, ease of maintenance, and product fit.
At Storm Reply, our approach is to pair the edge device capabilities of AWS IoT Greengrass with the storage and processing power of AWS and pairing these tools with a custom, high-quality web application. This approach is easy to maintain, flexible, stable, and scalable enough for the high standards and expectations of IIoT customers.
If you need support with building flexible, scalable, serverless, and modular IoT platforms powered by AWS, please contact Storm Reply.
Reply – AWS Partner Spotlight
Reply is an AWS Premier Tier Services Partner and MSP that specializes in the design and implementation of solutions based on new communication channels and digital media.
*Already worked with Reply? Rate the Partner
*To review an AWS Partner, you must be a customer that has worked with them directly on a project.