In this blog post, we cover some of the use case scenarios that are most suitable for deploying an edge computing platform.
The concept of edge computing is an extension to the cloud computing. With more and more applications deployed on the cloud, interesting patterns emerge in terms of usage and traffic flows. A more detailed analysis of these patterns leads to clusters, either in the form of user or data hotspots or in the form of common services being accessed too frequently. Therefore, cloud-hosted applications experience huge traffic coming from a bunch of users in a closed geolocation area. Or you may experience a spike in traffic to a specific API service or endpoint.
When that happens, it has an undesirable effect on the service performance. Chances are that it will also impact the network bandwidth and push the API performance statistics towards the red zone. An edge computing platform eases these problems, by distributing the load such that the clusters are served or accessed locally. Of course, every edge computing platform also relies on a centralized application hosted on the cloud. Still, the bulk of the processing and data orchestration happens at the edge. Here is a short article by PTC about the essence of Edge computing as "Meeting things where they live."
In this post, we cover a list of the five major use case categories that are the biggest beneficiaries of an edge computing platform. While most of these categories are, in reality, an individual industry segments, there are some categories which are in the form of usage trends.
Large manufacturing units and factories are always a source of cluster for man-machine interaction. Moreover, most of these establishments run mission-critical processes where real-time response to incidents and anomalies is the critical requirement. A cloud-based application for managing machines can lead to additional latency in data transfer. In addition to that, the factory employs IIoT based solutions for predictive maintenance of machines. The volume of data generated by these applications is humongous, and the cloud application can get swamped.
Considering these inefficiencies with cloud-hosted IIoT applications, edge computing is most relevant. Not only does it help in improving the real-time response to events, an edge optimized predictive analytics is 100x faster than a cloud-based solution. With the sheer volume of machine data needed for proactive mitigation of risks in machine failure, an edge optimized IIoT platform is a must for large industrial setups.
The Industrial Edge platform by Siemens is a total edge computing solution for such application scenarios. It supports a complete suite of management solutions, including edge device management and hosting of edge applications that leverage RPA and Machine learning for predictive analytics.
2. Media Processing
Just like predictive analytics of machines, media processing also requires a lot of processing power and high bandwidth. Applications such as CCTV based video monitoring are always deployed as a high-density local cluster of cameras and DVRs, which produces and consumes the media at the same time. However, for additional processing during video analytics and monitoring, a round trip from the cloud becomes inefficient if it is done too often. Surveillance applications for securing very high-security zones and for guarding precious resources often require such advanced features for real-time media processing.
An edge optimized media processing platform is the only way to expedite the monitoring and capabilities of the entire surveillance system. AllGoVision offers a suite of AI-enabled solution for video monitoring and analytics. The AllGoVision Edge Analytics suite is capable of directly taking the camera feed and processes and stores the alert videos and images for further processing. It can perform the standard security analysis procedures at the edge, such as intrusion detection, suspicious incidents, and counting.
Applications that depend on mobility or commuting are a constant challenge for network providers. Autonomous or self-driving cars, retail joints, public events, commercial places. All these are hot spots for user activity. In the case of autonomous cars, it is a machine to machine communication as there is much information shared between the cars. This creates a local hot spots across the roads and intersections. For retail stores, there is an increasing trend towards engaging with the customers inside the store. All of this calls for edge optimized online services.
For vehicles on the road, the wireless access network is optimized for V2X (Vehicle to Everything) communication. It provides the necessary edge optimized packet routing. 3GPP, the governing body for the 4G and 5G cellular wireless, has framed one set of specifications. For 5G, 5GAA is working towards a standard specification that is interoperable across all automotive companies and telecom network vendors.
For the retail industry, an edge optimized local network empowers store managers to provide enhanced customer experience by ensuring efficient and smooth handling of operations at the back office.
Healthcare facilities rely on mission-critical systems to deliver medical services. Many interlinked systems are involved here and look after different aspects such as hospital management, patient care, electronic health records. Many other systems connect medical equipments, which are all tied together digitally to provide 24/7 service. This setup is the perfect foil for localized edge optimized system.
GE Healthcare's Edison platform offers an integrated suite of applications that assimilate data from disparate sources, and apply analytics or advanced algorithms to generate clinical, operational, and financial insights. The platform is hosted through cloud, but edge optimized applications are directly embedded in the devices within hospital premises.
Web-based collaboration tools have been around since the early days of the Internet. However, there are special collaboration needs that arise in an enterprise setup. These are mostly localized collaborative applications that are part of the industrial workflow. As an example, shop floor supervision requires frequent interactions with technicians and workers. Similarly, other industries also need such solutions where a massive field deployment in place, such as oil and gas. Supervision, monitoring, and training for field staff is a routine activity. For such application scenarios, augmented collaborative experiences can elevate the motivation of the staff as well as the quality of the outcome.
PTC's ThingWorx platform, along with the Vuforia engine provides a collaborative AR/VR platform. This platform has the ability to mashup physical spaces, things, and devices with the digital medium to create an unparalleled virtual experience.
This technology helps not only in supervision and training but also in design reviews, shop floor audits, and inspections. In this case, the ThingWorx platform powers the edge network that connects the AR/VR platform with the devices and integrates device data into the virtual environment.