From monitoring to mastery: Enhancing industrial efficiency with advanced asset management

Problem statement 

Enterprises in industries such as mobile work machinery, manufacturing, transport and logistics, process automation, and energy production, are facing common challenges with globally spread assets and fragmented data, limited insights into performance, inefficient maintenance practices, and limitations in integration to business planning processes.

Without a comprehensive solution, organizations struggle to extract actionable intelligence from the vast amounts of data generated by their assets. This inefficiency leads to increased downtime, higher maintenance costs, and suboptimal operational performance. Moreover, as industries move towards service-based business models and remote operations, the need for robust asset management platforms becomes even more critical.

Enterprises require a centralized solution capable of providing, among others: 

• Real-time monitoring 

• Anomaly detection and predictive maintenance and

• Performance optimization 

• Seamless integration with existing systems such as ERP 

Asset-Fleet SYNC 

To address these challenges, we are developing an asset management platform designed to modernize how organizations monitor, control, and analyse their assets and leverage their data. Our platform, that we call Asset-Fleet SYNC, aims to address the complexities of modern asset management by offering comprehensive configuration and monitoring tools, analytics functionalities, and intelligence capabilities. In addition, the platform serves as a foundation for growing our expertise and gaining valuable insight to help our customers succeed. 

Through intelligent data collection, analysis, and visualization, our platform enables enterprises to make data-driven decisions, raise operational efficiency, and enhance overall business performance. Whether optimizing maintenance schedules, detecting anomalies in asset behaviour, analysing and reporting at any level of the infrastructure, or providing situational awareness and business intelligence, our solution aims to provide the agility and scalability required in today's competitive landscape. 

Solution overview 

Our solution is built on an AWS-based system architecture to meet scalability, reliability, and flexibility for any up-coming requirements. The UI is a traditional localized React frontend. Supporting it, is a double .NET8 backend system, with one instance managing static and configuration data, while the other handles heavier loads coming from edge locations' data collection and processing. 

The web application serves as a central hub for: 

• Administering the Identity and Access Management (IAM), 

• Reviewing of data streams originating from the edge  

• Configuring re-usable asset models derived from these data streams 

• Instantiating assets from the models 

• Visualizing of these assets as they stream data, providing with real-time insights into asset performance and operational status. 

Our development process benefits from the extensive expertise within our company's internal expert network. Drawing from past experiences and on-going projects, we capitalize on existing knowledge to inform our design decisions and ensure the robustness and effectiveness of the solution. Part of this robustness goal, for example, led us to additionally study an equivalent Azure infrastructure, ensuring a comprehensive understanding of alternative cloud environments and their potential impact on our solution's architecture, yet also offering more flexibility on our side to answer customer’s requirements. 

Technical Architecture 

The following is a detailed overview of the platform architecture and an elaboration of the various components. The picture below places the edge infrastructure in an imaginary factory, but it could also be any site, fleet, or device on land, sea, or air.  

Edge location - Source: This part of the architecture encompasses the diverse range of data sources within the factory environment, including sensors, PLCs, industrial machines, SCADA systems, vehicles, QA cameras, and CCTV. These devices generate both data and video streams, forming the foundation of the information flow within the system. 

Edge location - Collection: AWS IoT Greengrass acts as a gateway, communication broker, and local computing platform, it ensures efficient data collection from edge devices while also providing local processing capabilities. 

Data ingestion and streaming: Data from the edge location is ingested and streamed using a combination of AWS IoT Sitewise, IoT Core, and Kinesis services. IoT Sitewise facilitates the ingestion of industrial data, IoT Core handles communication with edge devices, and Kinesis enables real-time data streaming, ensuring smooth and reliable transfer of data to downstream processing pipelines. 

Aggregation, analysis, and storage: In this phase, data undergoes aggregation, analysis, and storage for further processing and insights generation. Kinesis Firehose is used for data delivery to various storage solutions, including Amazon S3 for scalable object storage and Amazon Timestream for efficient time-series data storage. Additionally, SageMaker, Amazon's machine learning service, is employed for advanced analytics and predictive modelling tasks. 

Monitoring, actions, and governance: The architecture covers a suite of AWS services for monitoring, triggering actions, and maintaining governance over the system. IoT Analytics and IoT Events enable the analysis of real-time IoT data and allow a quick detection and response to events. SNS (Simple Notification Service) eases communication and alerting. Lambda functions are used for serverless computing tasks. QuickSight provides an alternative data visualization through interactive dashboards for monitoring system performance. The web application serves as the interface for user interaction and control, ensuring seamless governance and management of the system. 

Key Features and Benefits 

The application covers the typical advantages of an AWS cloud infrastructure, ensuring high availability, durability, scalability, security, flexibility, and cost-effectiveness. With multi-AZ setup, failover mechanisms, read-replicas, load balancing, and auto-scaling capabilities, it guarantees uninterrupted access and performances. Managed AWS services, default cloud security measures, data encryption, and customizable firewall configurations ensure robust security, while frequent service updates and Infrastructure as Code (IaC) management enable high infrastructure flexibility. 

The Data API handles all data stream related functionalities, it is used both for the Webapp and for any customer’s own system integration. It offers flexible data processing options, including near/real-time streaming, data lake storage, and non-real-time processing, accommodating various data requirements seamlessly. Integrated IoT capabilities facilitated by AWS Greengrass ensure secure real-time communication, smooth pre-processing, analytics, and edge computing. Additionally, our platform integrates machine learning capabilities such as model training, deployment, tuning and monitoring, enabling data labelling, anomaly detection, enhancement of data insights and operational efficacy. 

 Roadmap 

At present, we have built the foundations of the product, including the implementation of the web application infrastructure utilizing Infrastructure as Code (IaC) with Terraform. We have set up an Identity and Access Management (IAM) system, with AWS Cognito integration. Our first focus has been on developing the Webapp Configurator, used to define the asset’s infrastructure and enabling the reception of data streams from Greengrass. Furthermore, we are adding basic monitoring features to enhance the platform's functionality. Finally, we have a functioning edge location simulation based on a windfarm, providing a tangible environment for testing and development purposes. 

Looking ahead, our roadmap outlines several key enhancements and additions to further augment the capabilities of the product to reach its goals. These include the integration of Kinesis for real-time data processing, refinements to the Webapp Configurator to automate as much as possible of the manual work, and improvements to monitoring functionalities making it more readable and reactive to user usage.  
Our roadmap includes plans to incorporate support for video assets for machine vison and learning purposes and setting up a data lake for efficient data storage and retrieval. Finally, we are planning the integration of data analysis and intelligence functionalities using AWS SageMaker. 

Conclusion 

Our journey through the development of Asset-Fleet SYNC is highlighting our push toward the crafting of innovative solutions to address the diverse needs of our customers. This time, leveraging both existing expertise and a freshly designed AWS-based system architecture, this project aims to offer intuitive interfaces, IoT stream data handling capabilities, and a scalable system for machine learning. With a solid foundation laid, our development process is moving smoothly forward while still committing in the continuous improvement of already realised parts. 

For further insights and information, don't hesitate to reach out to us!