Taking Computing to the EDGE
Dec 05, 2019
The Internet of Things (IoT) refers to the billions of physical devices worldwide that all connect to the Internet to share and collect data. Almost any physical object can be transformed into an IoT device, and more and more examples are mounting all around us in our everyday life. Anything from a light bulb that switches on remotely via a smartphone app, to quickly approaching autonomous vehicles are considered IoT devices. The number of IoT devices is believed to have reached 10 billion in 2018. That is higher than the number of mobile devices worldwide. Estimations vary, but Gartner predicts 25 billion IoT devices by 2021. With the Internet of Things rapidly expanding, it is imperative to understand edge computing and its role within the IoT development.
Edge computing is based around the idea that with more data than ever being collected and shared by IoT devices, computing and network infrastructure will necessitate a rebuild. Instead of data being stored in the cloud, a centralized third-party center, it will need to be analyzed and processed at the “edge” of the network, closest to the source creating it. With processing being done closer, devices will be able to deliver better performance, efficiency, and ultimately lower operational costs. But, with edge computing’s large upside and benefits comes potential risk. Adding edge computing devices that generate data to a network can lead to additional cyber security headaches.
First, let’s walk through some of the advantages edge computing offers:
- Speed: Moving applications for data analysis closer to the source of data generation reduces the physical distance and time required to move the data. By being closer to the data, edge computing can reduce periods of inactivity or delay and increase the responsiveness, speed, and quality of service.
- Reliable Connections: Because edge computing offers smaller, local data centers for storing and processing data, connectivity to IoT devices is much more reliable, in comparison to cloud services. For example, if an autonomous vehicle loses connection to the cloud, for even a second, due to a variety of reasons (inclimate weather, cloud server issues) it could have devastating effects. With edge computing autonomous vehicles have individual connections closer to the source of each vehicle, therefore providing faster reactions and more reliable connectivity.
- Lower Costs: Companies would be able to lower costs dramatically by decreasing the amount of bandwidth needed, replacing larger data centers with localized storage, and reducing overall data storage requirements.
All of the benefits above do come at a potential cost, security. Edge computing security remains a challenge due to the diversity of IoT devices and the fact that IoT devices do not have traditional IT hardware protocol. Edge computing’s decentralization of data gives it an advantage over the cloud, but ironically is also its biggest flaw and risk factor. Organizations that use the cloud store all their data in a centralized location and know that their data is protected by their cloud provider, by supplying the appropriate security tools. Edge computing faces security challenges, because devices are mostly small, individually operated, and not built with security in mind. These devices may not even be able to receive updates to help prevent attacks in the future. There is also a potential threat of cyber attacks that could target a single device as an easy entry point to gain access into the core system.
As the number of IoT devices continues to exponentially increase, it seems that edge computing could eventually be the next evolution of cloud-based data storage. It is imperative that organizations take time and strategize before implementing edge computing instead of surging to market for capital gain. Ultimately, if an organization wants to deploy edge computing, it's not the sort of thing that should be rushed or taken lightly: if you implement it cheaply, you could find that your devices are full of security holes that compromise your core system and might not be possible to fix.