Introducing the Intel® IoT Unified Edge Framework

Using workload consolidation to efficiently scale Industrial IoT solutions

Introducing the Intel® IoT Unified Edge Framework using workload consolidation to efficiently scale Industrial IoT solutions.

Executive Summary

The digital transformation known as the Industrial Internet of Things (IIoT) is creating new and complex challenges for many industries. Fragmented solutions, limited standards, immature security models, and inadequate approaches to the maintenance of digital assets are among the barriers that may prevent organizations from scaling valuable IIoT solutions. Alignment and/or convergence of the information technology (IT) and operational technology (OT) organizations is also an increasingly critical success factor.

Considering the growing number of edge devices, ever-increasing volumes of data, and existing infrastructure constraints, it is especially critical to reduce complexity and simplify scaling requirements. Two key concepts for achieving those goals are workload consolidation and orchestration. By combining those two concepts and defining a unified edge infrastructure, organizations can enable cost-effective scaling of IIoT and reduce capital expenses while improving security. Intel has made that easier by providing the Intel® IoT Unified Edge Framework, which is designed to help organizations consolidate workloads at the edge and efficiently scale their IIoT solutions.

This article provides more details about the Intel® IoT Unified Edge Framework, and presents a case study that explains how an industry-leading enterprise, Georgia- Pacific (GP), applied this innovative approach and the results the company achieved.

Areas of Opportunity

IIoT is transforming the way OT works by increasing efficiency, improving reliability, and reducing waste. The cloud is a powerful tool for many OT workloads, but it can fall short at times for IIoT use cases. Issues such as latency for control plane, massive volumes of remote data, and data aggregation (while maintaining context) require a complementing capability. The continuing evolution of edge computing is providing that complement and enabling the resolution of these issues. Yet, there are other key challenges at the edge that must be addressed:

  • Edge stacks are increasing in complexity. Due to lack of manageability standards, the management of these systems often requires manual intervention. This is frequently handled internally or through third parties, which leads to non- standard system approaches that drive additional cost and complexity while reducing reliability.
  • IT/OT convergence is a key challenge as well. In the past, keeping OT systems as a closed circuit with no IT integration was a common practice. That is no longer an option. In today’s competitive landscape, duplicating infrastructure in that manner is not an ideal solution.

Intel, in partnership with several Industrial Top 100 companies and with the goal of addressing these issues, established the Intel® IoT Unified Edge Framework, which is composed of a list of disciplines and guidelines that any customer may use to deploy a scalable and manageable IoT solution that provides real business value.

The Intel® IoT Unified Edge Framework provides a foundation for building end-to end IoT solutions. It includes guidance for selecting the right building blocks to support business needs and providing consistency, scalability, and completeness of IoT solution architectures across the enterprise. The framework also includes a reference architecture that customers may implement using their preferred building blocks.

Ultimately, this framework from Intel helps customers gain a clear understanding of the value of standardization in each of the disciplines it includes and the options available to them, so they can make informed decisions about what is best for their business.

The Value of Workload Consolidation

An important discipline for this modernization is workload consolidation. Consolidating new and legacy systems in edge computing environments can serve multiple purposes and lower the total cost of ownership in a number of ways. Examples include the following:

  • By leveraging a platform that can enable the orchestration of workloads, organizations can reduce infrastructure costs, make systems functionally safe and secure, and simplify system management.
  • Reducing the number of components enables organizations to streamline operations, improve productivity, and reduce cost and complexity.
  • Organizations can lower both capital expenditures (CAPEX) and operating expenses (OPEX) by significantly reducing the number of unique devices that must be kept on hand for maintenance and cutting the related costs of training and supporting staff.
  • Reducing costs associated with system obsolescence is another plus of workload consolidation. By consolidating workloads so that one device is hosting multiple use cases, organizations can ensure they will have fewer devices to upgrade or replace in the future.

GP is partnering with Intel on this initiative as one of the early pioneers to enable internal digital transformation.

There are two main components of the Intel® IoT Unified Edge Framework, shown as a multi-tier system ontology in Figure 2 as part of the edge-compute infrastructure:

  • Edge Compute Devices: These are the devices commonly referred to as IoT gateways. They have enough processing power to host several IoT solutions (each running systems based on Intel® Core™ i7 processors). Edge Compute Devices are deployed close to sensors, receive data directly from them, and also interact with actuators to perform specific actions. Edge Compute Devices are standardized by IT and include both hardware and software. Essentially, IT selects a device with certain hardware components, and the device includes the Ubuntu* operating system (OS) and a set of other software that enables it to run one or more use cases.
  • Edge Server: On-premise, high-processing servers (using Intel® Xeon® processors) are required for consolidating data coming from many Edge Compute Devices or for processing heavy workloads such as video analytics. These local servers provide independence from cloud providers, allowing higher availability, lower latency, and reduced data transmission.

Workload Consolidation Benefits

As described in the previous section, the main goal of workload consolidation is to reduce the total cost of ownership and enable the following benefits:

  • Decrease System Equipment Footprint: Instead of having a dedicated edge compute device for each point solution, a single, more resourceful device can host several IoT solutions/components.
  • Ease of Deployment and Management: By reducing the number of devices used by point solutions, organizations can considerably reduce the number of devices they need to manage, which means less operational work.
  • Increase Security: Organizations have the potential to minimize the attack surface of their network by reducing their hardware (HW), firmware (FW), and software (SW).
  • Reduce System Integration Complexity and Access to Data: A framework is required to allow different workloads to coexist in the same device. This simplifies solution deployment and integration. The enterprise data bus receives data provided by different solutions.
  • Improve Reliability of Underlying Process Control Systems: Minimize duplication of data requests and reduce the load on critical (but aging) control systems.
  • Ensure No Vendor Lock-In: Customers are able to get their data from each solution, and vendors have to adapt to HW defined by the customer.
  • Optimize Utilization of the Aggregated Compute at the Edge: Eliminate the inefficient use of edge-device resources, which only run a few services per device.
  • Accelerate the Adoption of IoT Technologies: Help enterprises with digital transformation.

Case Study – Georgia-Pacific

Georgia-Pacific is one of the world’s leading makers of tissue, pulp, paper, packaging, building products, and related chemicals. GP is aggressively investing in digital technologies to fuel transformation initiatives and continually improve production efficiency, sustainability, worker safety, and product quality in more than 150 manufacturing locations that the company operates across North America.

Advances in technology have enabled a wide array of IIoT solutions that, when applied in the right areas, can transform manufacturing operations. To remain competitive, however, organizations must be able to scale these solutions quickly and cost-effectively. Yet, the rapid adoption of IIoT solutions across industries has led to market fragmentation and a variety of technology options. Many vendor solutions deploy a full end-to-end stack, complete with proprietary edge-compute devices, resulting in installations operating a number of disparate, closed solutions, each requiring digital maintenance, integration, security, and support.

Although initially pleased with their IIoT solutions, Georgia Pacific found that this “silo” approach made it difficult to scale solutions when they sought to expand their IIoT capabilities. As GP discovered, supporting a series of disparate IIoT solutions becomes increasingly expensive and labor intensive over time. Long-term support costs add up quickly and undermine the primary drivers for implementing the solution. The increased complexity of the resulting solution can also reduce reliability and cause instability in source systems, such as manufacturing control systems.

With Intel’s help, GP learned that consolidating workloads in more resourceful edge-compute devices, and encapsulating computational services using containers and virtual machines, enabled the company to manage and deploy suites of IIoT solutions at scale.

Attempting to scale IIoT solutions without consolidating hardware and using a more efficient approach for managing workloads can become daunting. Maintenance and support of computing assets for a large manufacturing enterprise is already an enormous undertaking. Adding more nodes that require a high level of uptime and performance makes the task even more difficult.

Based on the experience of deploying the Intel® IoT Unified Edge Framework at the GP Muskogee facility, the combined Intel and GP team identified how leveraging the framework could dramatically reduce the number of resources (both labor and hardware) required to deploy and maintain IIoT systems and enable scaling.

The local teams were able to consolidate compute workloads for three disparate IIoT solutions from three different providers into one standardized compute stack. The solutions required varying degrees of I/O from the field gateway and compute from the edge server.

The consolidated workloads listed below are representative of typical IIoT solutions currently being deployed across the enterprise.

  • Computer Vision-Based Anomaly Detection Service: This system uses fixed cameras to identify changes in the operating environment. Examples of how this capability can be applied include tracking materials to make sure they are not placed in unauthorized areas, which could lead to safety hazards such as blind spots or exit obstruction; detecting whether the shielding that protects machinery has been damaged; and identifying whether a machine is operating in a typical or atypical fashion, to proactively identify events that may affect quality and reliability.
  • AI-Based Object Detection Safety System: Reduces safety risks in a variety of ways – for example, by using AI to detect pedestrians in high-traffic areas for mobile equipment, and providing greater visibility through lighting and/or automated gates.
  • Long-Range Environment Sensor System: Retrofit sensors and connect plant equipment that was previously cost prohibitive to add to the network. Because these assets are often outside the main production areas, information from these sensors can provide additional visibility into operating parameters that may lead to improved reliability and efficiency. These three workloads are a small representation of related capabilities that, when applied in combination, can enable the transformation of manufacturing operations and lead to improvements in production efficiency, sustainability, worker safety, and product quality.

Based on knowledge gained from the initial deployment, the team has estimated the following savings compared to the legacy approach:

Expected Savings Cost Reduction Contributing Factors and Notes
Maintenance and Support 30%+ Reduced training and troubleshooting; simplified integration, patching, upgrades, etc.
Hardware 30%+ Initial cost with 5-7 year replacement
Performance and Uptime 10% Reduced MTBF and MTTR (labor only)
Initial Deployment (One Time) 40%+ Image and workload deployment, networking, mounting, activation

All of the devices required to support these workloads are managed by IT as standard devices, with regular manageability tools, having secure standard OS images for the edge-compute devices and a virtualization environment on the edge server side. This helps with IT/OT convergence.

Regarding data management, there is a MQTT (Message Queuing Telemetry Transport) broker per tier, which enables GP to capture information from IoT solutions, removing vendor lock-in. This allows the company to produce the aggregated/enriched data required for insight generation.

Also, by having a platform that can help with the orchestration of workloads, GP can efficiently manage its infrastructure resources by deploying workloads on the best available nodes at a certain time.

The Intel® IoT Unified Edge Framework is clearly helping Georgia-Pacific on its digital transformation journey. By using the Intel framework, GP is accelerating its deployment of new IIoT solutions in a more sustainable way, and simplifying the scaling up of these solutions across its operating facilities while minimizing the costs of running and maintaining this new platform.

Conclusion

As the Georgia-Pacific case clearly shows, the Intel® IoT Unified Edge Framework provides the disciplines and guidance that organizations need to create and efficiently scale end-to-end IIoT solutions. By using the Intel framework to consolidate and orchestrate workloads at the edge rather than continuing to support an increasingly complex and costly array of disparate IIoT solutions, organizations can improve security, performance, and reliability while reducing capital expenses and lowering their total cost of ownership.

Authors

Dave Nettuno
Georgia-Pacific
Enterprise IoT Architect

Kit Fennell
Georgia-Pacific
Technology Leader

Dalibor Labudovic
Georgia-Pacific
System Engineer

Marcos E. Carranza
Intel Corporation
Senior IoT Solutions Architect

Cesar Martinez Spessot
Intel Corporation
Engineering Director
Senior IoT Solutions Architect

Lakshmi Talluru
Intel Corporation
Sr. Director Digital Transformation

Jennifer Frieda
Intel Corporation
Account Executive