Accelerate operations excellence with 5G private networks, edge computing and IoT

Manufacturers’ see an increased need for flexibility and cost-effectiveness in running their factories, driven by global competition and changing customer demand globally. This has led to a significant push for digital transformation (Industry 4.0) to identify and address inefficiencies more proactively throughout the system.

Enhancing operational agility and flexibility is one of the top three actions of CEOs to build a change-ready organization. This produces (i). flexibility of production practices, (ii). need for real-time decision making on the floor, and (iii). the ability to contextualize plant information for minimizing downtime, as well as the key success factors of operations.

The CIO/CTO needs to make sure of the scalability of systems and infrastructure used on the shop floor that control bandwidth and storage, as well as minimize repeatable costs. This becomes a challenge with a shortage of trained IT resources in manufacturing.
The Operations Director needs to interpret diverse data sets in disparate data stores (shift records, SAP, process historian, etc.) to root daily operational issues. At the same time, they need to review production end-to-end to monitor process, as well as Overall Equipment Effectiveness (OEE) both locally and globally, to minimize downtime.
The Line/Shift Manager is tasked with continuously optimizing resources and processes on the shop floor while keeping workers informed about metrics. They are responsible for identifying and addressing potential incidents on the shop floor to maximize OEE, and assigning the appropriate resource to support tasks.
Service technicians must identify and resolve equipment issues quickly. They need access to guidance and holistic asset performance insight, such as maintenance schedules and history, and engineering notes.
The Plant Manager needs to keep processes optimized at all levels. For this they must be supported through closed loop low latency OEE monitoring for low-latency monitoring of critical processes even in offline environments. At the same time they must able to use and contextualize the monitoring data to make operational decisions at the factory level (maintenance, planning, scheduling, etc.).

We outline seven cases that use the intersection of Industrial IoT, 5G Mobile Private Networks (MPNs), and Edge and Cloud Computing, within industry 4.0.

In industrial settings, common challenges include 1/ connecting sensors to brownfield equipment and configuring contextualized data streams, 2/ extracting equipment telemetry from one or more edge sites and storing on the cloud for global OEE monitoring, 3/ being able to monitor local equipment telemetry for connection independent operations, 4/ providing familiarization with the environment, 5/ being able to see and process data using augmented reality (AR), 6/ preventing failures by leveraging anomaly detection services, and finally 7/ using computer vision at the edge to monitor defect rates.

In this post, we present the Industrial IoT (IIoT) 5G demonstrator labs, a joint solution co-developed by AWS and Partners – Vodafone, IFM, Matterport, and Treedis. The IIoT 5G demonstrator aims to illustrate the value of gathering and processing industrial data, showcasing the use cases that resonate most with our respective customers. It also shows how our reference architecture supports these use cases, and building a foundational data platform to allow customers to evolve and expand their digital transformation.

In order for industrial customers to unlock value from their data, they need a coherent data strategy that includes a data management and connectivity architecture.

“Industrial Transformation efforts need successful, proven partnerships that accelerate time to value. Through the industrial transformation labs initiative, we demonstrate how multiple specialist partnerships seamlessly enable end-to-end use cases through the convergence of technologies such as IoT, 5G, artificial intelligence/machine learning (AI/ML), and spatial computing. The lab’s initiative enables customers to quickly understand not only the benefit of the technologies but also how they fit together. This means customers can deploy complex initiatives and start to see return on investment faster .” – Bryan O’Flaherty Wills, WW GTM Lead, Industrial IoT, AWS

Architecture

Figure one: High-level architecture for the end to end use cases at the lab

The IIoT 5G demonstrator solution has the following components:

IFM sensors, cameras (details in the following sections), use of IO-LINK masters and IFM Moneo applications with direct connector to AWS IoT core and as an OPC-UA server.
AWS applications using AWS IoT SiteWise, Amazon Lookout for Equipment, AWS IoT TwinMaker, AWS IoT Core, Amazon Athena with Amazon Simple Storage Service (Amazon S3), and at the Edge (factory) with AWS IoT Greengrass and AWS IoT SiteWise Edge.
Vodafone offers the Vodafone MPN with Indoor Coverage, which uses two Radio DOTs.
Vodafone Edge Gateway to connect the assets and IoT sensors to 5G MPN that runs AWS IoT Greengrass.
Matterport 3D capture that provides the immersive spatial contextualization of a Digital Twin, embedding the tags and labels of the equipment and factory data or demo space floor (such as motors and sensors).
Augmented Reality (AR) and virtual reality (VR) workflows by Treedis on top of the Digital Twin for remote inspections and virtual training on Standard Operating Procedures.

Architecture overview

AWS IoT SiteWise and AWS IoT TwinMaker are two industrial-focused services from AWS that work together in this architecture to support manufacturers in their digital transformation journey. These services leverage the power of the Internet of Things (IoT) and cloud computing to streamline operations, enhance efficiency, and drive innovation.

AWS IoT SiteWise offers the unified data foundation to collect, organize, and analyze industrial data at scale with extended industrial protocol support. Its asset modelling and data stream management features allow users to add context to sensor or process data captured at the industrial edge and associate those measurements and state updates with the assets to which they relate. In addition, SiteWise offers analytics features that enable users to define transformations and metrics at the model level, which may then be applied to multiple assets associated with those models. Typical examples of transformations and metrics would be unit conversion and rolling averages, maximums and minimums for various periods, thereby allowing users to standardize data and easily draw performance comparisons across multiple sites.

With AWS IoT SiteWise Edge users can Connect Legacy OT systems to the cloud with edge gateways. Data sources could involve many different new or legacy protocols, and SiteWise Edge supports collection and translation across these protocols through built-in partner integrations. Data can be captured directly from OPC-UA servers. Just like SiteWise in the cloud, SiteWise Edge allows users to run local data processing to calculate metrics at the edge and use this to drive local dashboards and alerting. In this reference architecture, connectivity is provided by a local private 5G network, both for gateway data backhaul, IP based sensors such as inspection cameras, and user experiences for factory workers.

IFM sensors and cameras (more details in the following section), and IO-LINK master eliminate the need for expert skills needed to quickly install, register, and capture data using industrial gateways, with sensors that conform to the IO-Link standard, configured using a managed solution such as Moneo. These gateways are connected using private 5G to avoid the management of multiple WIFI access points across a large industrial site.

A recent feature added to SiteWise in the cloud is native integration with Lookout for Equipment. Once again, through predictor definitions specified in reusable models, SiteWise gathers chosen sensor signals and makes them available as training data for anomaly detection models in Lookout for Equipment. Through the SiteWise console or APIs, users can schedule inference jobs that run at regular intervals and report anomalies along with indicators about the signals that are primarily responsible, giving operations teams valuable insights for root cause analysis. The advanced warning of equipment failure before plant outages allows for maintenance schedules to be optimized so that teams can tackle similar jobs efficiently, making sure the right parts and skills are available, and context-switching is minimized.

With suitable product inspection cameras in place, images of good and bad products can be used to train ML models using Amazon Lookout for Vision. Through the use of AWS IoT Greengrass components and AWS IoT Core jobs, Lookout for Vision can deploy these defect detection models to image inference processors running at the edge. As products change over time or more defect information is captured, these models can be retrained and redeployed to the edge remotely.

AWS IoT TwinMaker is used to visualize centralized data and metrics in the context of Matterport 3D point-cloud scans, providing a full digital twin of the global operation. A key feature of TwinMaker is to add extra context not available from the plant floor. This could be work order history or maintenance instructions, which may be indexed in ways that are hard to associate to the related sensor data. Through the use of a knowledge graph, TwinMaker users can make these new associations between disparate information sources to make them available in the digital twin at runtime. Matterport and AWS IoT TwinMaker are deeply integrated to provide an immersive 3D experience that allows remote or unfamiliar workers to comprehend a facility from a distance, yet still have awareness of time-related machine telemetry and state data. By layering the digital twin with the Treedis Cross device XR workflows, local workers can instantly gain deep data insight into the state of a machine or production line while viewing it in the real world. AWS IoT TwinMaker provides the tools to construct these immersive 3D experiences, also called scenes, along with unified data services to query the knowledge graph and time series data. TwinMaker offers several ways to embed these scenes and data in Grafana, the popular dashboarding tool, third-party partner products, or to embed them in your own company web and mobile applications.

In the next section, we detail how different capabilities flow together to create an end-to-end IIoT 5G demonstrator solution and multiple Line of Business applications.

1. Bringing data from the Edge to AWS Cloud
The IFM group of companies is a global industry leader for innovative sensors, controllers, and ERP-based solutions for supply chain management and shop floor integration worldwide. Continuous evaluation of process data is the best basis for sustainably successful business decisions. In order to obtain the important information from the shop floor, the reliable connection of the sensors to the IT infrastructure is essential. IFM Moneo, an IIoT platform that combines the level of operations technology with the level of information technology. The sensor data generated in the production plants can be read, processed easily, and used as a basis for sustainable corporate decisions.

Moneo has a modular structure and consists of basic software as well as applications, such as for condition monitoring or for IO-Link sensor parameter setting. This makes it possible to put together a tailor-made software package for every individual requirement.

Moneo’s modular concept also provides a selection of different, easy-to-handle applications that can be linked together. From the sensor parameter setting there is the automatic onboarding of sensor level data, as well as the immediate realization of sensor values and edge connectors that enables the outgoing data transfer to AWS IoT SiteWise through either the MQTT or OPC UA server.

Simple: Completely pre-installed to be set up without IT expertise in the manufacturing environment, production network, or data centre
Convenient: Integrated functions for IT maintenance (updates, back-up, and support)
Safe: Robust hardware with standard interfaces and pre-installed software for increased process and operational reliability, in two versions
Central: Network node for data collection and processing

2. Connectivity considerations: when Private Networks matter
Vodafone’s MPN connectivity offers future proof, secure and robust private 4G and 5G connectivity to enable incremental value through use-case overlays that address safety, efficiency, visibility and growth priorities. It consists of: dedicated radio network, dedicated core network deployed on customer premises using dedicated spectrum allocated by the local operator. An MPN supports mission/business critical data communication of the customer according to one or more of the following characteristics: Improved coverage and reliability, Ultra-high performances, Security. Vodafone provides Edge Computing services along with MPN to deliver a full end-to-end secure service. Edge Computing processes the data close to the point of origin or consumption, which means you can add intelligence to much of the equipment that sits within operations, and apply AI-powered automation at those end points.

Industries don’t always want all their data shared to the cloud, due to sovereignty, privacy, and commercial reasons. Vodafone’s managed Cloud, Edge and MPN services can support customers to securely connect with the Cloud.

Recognising that customers require an end-to-end solution for their business, Vodafone’s 5G Edge Innovation Lab in MediaCity, Manchester was the perfect location to bring Vodafone, AWS, and Partner technologies to life with live demos and pilots. We used this location to enable the AWS ecosystem of industrial partners to showcase a combined live solution to engage with customers. Together with the partners we created an end-to-end solution reference architecture that integrates Vodafone’s MPN 5G SA, Edge, IFM’s IoT Sensors, AWS Edge to Cloud Services, the Matterport scan, and the Treedis solution.

Key considerations Vodafone addressed during the end-to-end solution design and deployment included:

Selecting the most appropriate MPN and devices for a business be that proof of concept/trial MPN, dedicated MPN, local hybrid MPN, global hybrid MPN, network slice.
Sizing the appropriate Dedicated Edge Compute for the uses cases and pairing with an appropriate 5G router to support 5G Standalone (5G SA).
Deploying Cloud services to the chosen Edge Compute hardware in conjunction with OT partners, making sure that translation, collection, and sharing of data to the Cloud is suitable.

Figure two: Vodafone connectivity options

3. Extract equipment telemetry from the Edge and storing on AWS for OEE monitoring

To extract equipment telemetry from industrial equipment in order to aggregate OEE and other critical metrics across multiple sites due to locally hosted historians, this joint offering leverages the AWS IoT SiteWise fully-managed service to standardize data models across all sites, collecting data into a central, cloud-based repository. This approach not only facilitates the automatic calculation of vital metrics, but also enables the sharing of these metrics and insights through cloud-hosted dashboards.

The use of AWS IoT TwinMaker for visualizing centralized data and metrics within the context of the Matterport detailed 3D point-cloud scans provides a comprehensive digital twin of global operations. This allows for an unprecedented level of insight and control over manufacturing processes. For operations executives looking to dive deeper into local production events remotely, Amazon Kinesis Video Streams offers the capability to stream live or historical video directly to the digital twin. This feature enables the navigation of video streams from one or more sites using the visually rich Matterport rendered twin. This makes it easier than ever to monitor, contextualize, and share data and live video for individual sites or an entire enterprise with stakeholders across the organization.

4. Remote viewing options through Digital + VR and on-site access with AR

The narrative of immersive technology integration continues to unfold with remote viewing options through Digital, VR, and on-site interactions with AR provided by Treedis. These platforms extend a kaleidoscopic lens to your operation, going beyond the constraints of physical boundaries, and offering a comprehensive view into your operations.

This all starts by capturing your facilities and the creation of digital twins – accurate digital representations of physical environments, powered by the Matterport digital twin platform and seamlessly integrated with AWS IoT TwinMaker. This integration enables facilities managers to unlock unprecedented levels of operational efficiency, streamlining the management of complex data sets, and making sure of the consistent high performance of facility assets. By analyzing trends and predicting maintenance needs, equipment uptime is increased significantly, especially in remote facilities, and field operations in manufacturing plants are improved through real-time IoT sensor and process data.

Figure three: Matterport’s space capture integration with AWS IoT

Moreover, this solution champions flexibility and productivity by offering a bespoke, single, unified interface that allows easy access to all connected data sources within a Matterport digital twin. It simplifies the process of creating these digital representations, making it more efficient and cost-effective than ever before.

From an Environmental Health & Safety (EH&S) perspective, the Matterport and AWS IoT TwinMaker integration reduces the need for site visits, and consequently lowers the carbon footprint. This not only helps achieve sustainability goals, but also reduces operational costs and minimizes the risk of injury, aligning perfectly with the core objectives of modern facilities management.

With a unified 3D view of business operations, site managers can feed all essential application data into their Matterport digital twin, offering a comprehensive overview of operations. This facilitates remote collaboration, reduces expenses, and significantly increases operational efficiency by embedding live data into the digital twin.

By leveraging existing IoT, video, and enterprise application data without the need for reingestion or moving data to another location, the integration saves time and resources. An automatically generated knowledge graph binds data sources to virtual replicas of physical systems, accurately modelling real-world environments. This offers an immersive 3D view of systems and operations, optimizing efficiency, increasing production, and improving performance.

Treedis merges digital twin technology with extended reality (XR) to revolutionize physical spaces across augmented reality (AR), virtual reality (VR), and desktop platforms. AWS, Vodafone and Matterport and Treedis culminated in a pioneering IoT project: the creation of a live, interactive digital twin for the Vodafone Edge Innovation Lab.

This innovative solution permits field workers to access real-time data via AR, providing guidance and instructions needed for agile decision-making. This approach not only boosts the operational efficiency and safety of workers but also seamlessly integrates them into a smart, interconnected workplace environment.

Bringing the pieces together for the IIoT 5G demonstrator

The industry is looking for fast ways to adapt and drive the benefits of technology to reduce costs and innovate to gain competitive advances. This showcase has been brought to life from concept to reality in six months, with all partners actively collaborating, bringing together their best skills to create a combined service that can be leveraged across the factory plant/ shop floor:

Industrial IoT sensors for real-time monitoring, optimization, and predictive maintenance with Digital Twins to visualize and navigate over the factory without travelling 1000s miles, thereby also supporting sustainability
5G MPNs and Edge gives faster response times, improved scalability, security, reduced network congestion to optimize data intensive applications, and immersive user experiences
Provides manufacturers with secure data transmission, seamless collaboration, and adaptability to industry requirements
Brings the reduction of operational overhead with self-service onboarding, dashboards and alerts, predictive maintenance, and CXO reporting on OEE to manage multiple factories and countries with scale.

Conclusion

In this post, we presented the IIoT 5G demonstrator to illustrate the value of gathering and processing industrial data, showcasing the use cases that resonate most with our respective customers. It also shows how our reference architecture supports these use cases and builds a foundational data platform to allow customers to evolve and expand their digital transformation.

Jenn Didoni, Head of Cloud, Edge & MPN, Vodafone Business, says “By having a live demo that a customer can tangibly interact with, we are giving customers a chance to try before they buy. In collaboration with our ecosystem, we’ve created this demo enabling customers to experiment with new technology that touches the heart of their operations. Showcasing like this makes sure that it delivers the value and results needed, accelerating adoption in the long run.”

The demonstrator helps customers intuitively understand how to unlock value from their data with a cohesive data strategy that includes data management and connectivity architecture.