Adopt edge computing to accelerate the real-time decision-making process

We recently outlined five trends that are imminent in the IoT and Industry 4.0 landscapes, and edge computing is one of them. We predicted that edge analytics would become the new de facto enabler of near-real-time decision making. Edge computing addresses the challenge of a huge amount of data analysis by processing a significant chunk of the data locally and pushing the filtered data into the cloud. Industrial enterprises across industries are also inclined to leverage edge computing as it expedites the overall decision-making process. In this PoV, we describe the benefits of edge computing including recent examples of deployments and outline how an industrial enterprise can leverage the same for its decision-making process.

Edge computing offers a bouquet of benefits across both technology and business areas

As different enterprise business functions are connected through IoT devices, organizations are looking for real-time decision making. Cloud computing requires a longer time for decision making because all of the data must be loaded into the cloud before it is processed. Edge computing can resolve this challenge because the data is processed near the edge devices. Edge computing provides several benefits, as described in Exhibit 1.

Exhibit 1: Edge computing simplifies data processing with real-time business outcomes

Source: HFS Research, 2020

Industrial enterprises have started to deploy edge computing solutions in their landscape

Enterprises are targeting several benefits such as cost reduction, safety improvement, operational efficiency improvement, etc. through edge computing. Some of the examples include.

• Siemens’ Amberg plant is saving approximately 12,000 Euros in downtime costs per annum per equipped machine through edge computing
• Honeywell is leveraging edge computing to reduce in-flight cancellations, test flights, and scheduled maintenances, increasing the overall safety of helicopters
• GE Transportation’s Evolution Series Tier 4 Locomotive uses on-board edge computing to improve operational costs, safety, and uptime

Leading IoT platforms have already deployed edge analytics functionality

IoT platform providers have also started to built-in the edge analytics functionality as a part of their offerings. Some of the examples include:

• AWS offers several services for edge devices. AWS IoT Greengrass processes data locally and leverages the cloud for management, analytics, and durable storage. Amazon FreeRTOS is an open-source operating system for microcontrollers that makes small and low-power edge devices. AWS IoT Device Tester is a test automation tool for IoT devices that determine whether the device will run Amazon FreeRTOS or AWS IoT Greengrass and interoperate with AWS IoT services.
• Azure IoT Edge has three components: IoT Edge modules (containers that run Azure services, third-party services, or own code; these modules can be deployed to IoT Edge devices), IoT Edge runtime (runs on each IoT Edge device and manages the modules deployed to each device), and cloud-based interface (enables remote monitoring and management of IoT Edge devices).
• An integral part of the Predix Platform is Predix Edge, which enables asset connectivity and local intelligence for the IoT landscape, delivering connectivity, apps, and intelligence from the plant floor to the data center.

Five imperatives for enterprises to implement edge computing

The benefits of edge computing look promising, but enterprises need to focus on these pointers to reap its full benefit.

• Reprioritize decision making: Enterprises need to analyze the complexity and outcomes of business decisions. Based on the maturity and impact of decisions, enterprises need to leverage cloud data for strategic decisions, whereas edge computing can support the threshold decisions.
• Focus on process re-engineering: As some business applications will be used for automated decision making rather than simple data aggregation and integration, process re-engineering is necessary, based on the business requirements. Some industries and geographies have a strict rule about data privacy, data security, and documentation of process quality, etc., so enterprises must consider all the regulatory policies and compliances before the implementation.
• Revamp the governance framework: Organizations should focus on empowering line managers to take ownership of some of the decision-making processes. There must be clear guidelines for escalation and approval from central command for certain aspects of decision making.
• Manage the technology priorities: Enterprises need to consider data pre-processing, data volume, and data velocity, among other aspects, for edge processing. Also, enterprises need to manage and integrate diverse IoT gateways, protocols, and standards for edge computing.
• Attain the operational cost efficiency: Though edge computing reduces some of the operational cost (related to storage, network, etc.), enterprises need to invest in edge servers, local server cybersecurity, firmware upgrades, and other areas related to edge computing. Enterprises should perform a cost-benefit analysis before implementation.

The Bottom Line: Decentralize the decision-making process and leverage edge computing for the selected threads of overall decision making.

As most of the enterprise decision-making processes are still controlled from a central command center, business processes and overall governance are aligned accordingly. Enterprises need to decentralize the decision-making process and classify the threads of decision making that can be handled locally and the rest from a central location. Edge computing facilitates the local decision-making process, but enterprises need to focus on both business and technology aspects for effective implementation.