Cloud computing has been a crucial catalyst for the Internet of Things. Cloud platform technologies provide on-demand scalability for all kinds of IoT applications, enable the transformation of data into actionable information, and thereby create opportunities for new types of applications, business models and value propositions.
However, the skyrocketing number of connected devices is starting to challenge the prevailing cloud infrastructure. Furthermore, the cloud model is not really efficient when internet connectivity is poor, or when real-time data processing is of the essence.
Enter fog computing, a different architectural paradigm which extends cloud computing and services to the edge of the network, allowing high-speed data processing, analytics, decision-making and action to happen closer to the devices in the field.
What is fog computing?
Coined by Cisco, the term “fog computing” (or “fogging”) describes a decentralized computing infrastructure, where computing resources and application services are brought to the edge of the network, introducing an intermediate processing layer between IoT devices and the cloud. A fog is hence “a cloud close to the ground”:
An emerging wave of Internet deployments, most notably the Internet of Things (IoTs), requires mobility support and geo-distribution in addition to location awareness and low latency. We argue that a new platform is needed to meet these requirements; a platform we call Fog Computing, or, briefly, Fog, simply because the fog is a cloud close to the ground. 1
Whereas traditional cloud computing transports all data to a central server, fog computing enables analytics and other computing operations to be performed closer to the network’s end nodes – the “things” that produce and act on data. The objective is to perform computing operations at the most appropriate point along the chain between the data source and the cloud. For example, time-sensitive data will typically be handled close to where it is collected, while data for historical analysis and long-term storage is sent to the cloud.
While the term fog computing is sometimes used interchangeably with edge computing, edge computing usually implies that processing and communication take place at the device level itself. A fog computing architecture, on the other hand, typically includes a fog node or IoT gateway at the local area network level, which processes data received from multiple end points.
Benefits and use cases
The first obvious advantage of fog computing is improved network efficiency. By analyzing and acting upon IoT data closer to where it is collected, fog computing can eliminate the transportation of large volumes of data and significantly offload the core network.
“The growth in IoT is explosive, impressive – and unsustainable under current architectural approaches. … Fog computing adds a hierarchy of elements between the cloud and endpoint devices, and between devices and gateways, to meet these challenges in a high performance, open and interoperable way.” 2
Compared to traditional cloud computing, fog computing can also enhance service quality, thanks to greatly improved data rates and especially reduced service latency. Real-time or predictable latency is critical for many IoT applications, including various kinds of anomaly detection and automated alerts, vehicle-to-vehicle communication and e-health, and areas such as gaming and augmented reality.
Cisco advises to particularly consider fog computing in scenarios where:
- data is collected at the extreme edge, such as in vehicles or ships or on factory floors;
- thousands or millions of things across a large geographic area are generating data;
- it is necessary to analyze and act on the data in less than a second. 3
New business opportunities
Just like the cloud, the fog is poised to spark the development of new business models and new revenue opportunities. While corporations like Arm, Cisco and Intel are currently dominating the fog computing scene, there is also a growing number of smaller companies entering the arena with fresh ideas and solutions for handling and exploiting IoT data.
For example, Silicon Valley-based FogHorn Systems offers an application development and deployment platform designed to enable edge intelligence and analytics for industrial IoT solutions. Targeting industries such as energy, manufacturing, mining, transportation and healthcare, the platform processes data directly on distributed edge devices or sensors instead of sending all data to the cloud.
Similarly, French startup SpinalCom is developing a “micro-middleware” IoT platform called SpinalCore which enables its customers to create fog-based systems, connect incompatible devices, build new applications and deploy them fast.
Other companies are creating fog-based solutions for specific target industries or use cases, such as Marsec, which develops applications for optimizing marine operations, and Nebbiolo Technologies, which offers a computing platform for industrial automation.
Fog computing is still in its infancy. The concept isn’t proven or even fully developed, it lacks standardization, and much remains to be done in order to realize its potential and facilitate its practical adoption. But this also means there is a vast field of opportunity. For developers and companies with ambitions to capitalize on the growth of the Internet of Things, then fog computing might be a domain worth exploring.