What do a college student, a space probe, and an Internet of Things (IoT) controller in a refinery have in common? When it comes to making critical decisions, a sensor’s proximity to intelligent systems and data makes a huge difference in positive outcomes. Let’s face it, a lot of us went out-of-state for college to avoid such proximity, and space exploration guarantees long “latency” between a space vehicle (sensors) and ground control (intelligence) – round trip communication to Mars is not less than 13 minutes. The farther away these things go from their origin, the bigger the latency gap becomes.
Proximity for Industrial IoT Devices
In the “boomable” environment of the oil and gas industry, situational awareness is critical; which in turn is driving the deployment of Industrial IoT devices (IIoT). A single refinery can have 10,000 IIoT sensors across a plant, which itself can be longer than a mile. To make critical decisions in a refinery – the kind of decisions that affect safety, efficiency, and security – you need to have the right application with immediate access to critical data. In other words, decision making software can’t be orbiting Mars. In fact, it can’t even be miles away in a public cloud. The intelligent application must be very near to the physical device it’s controlling. IoT and software require proximity for critical decision making.
Industrial automation and control is rapidly driving the modernization of many process-driven industries such as Food and Beverage, Mining and Metals, Pharmaceutical, Utilities, Pulp and Paper, and many others. Leading the way is an industry group known as the Open Process Automation Forum (OPAF), in which CPLANE.ai and over a hundred others such as Exxon Mobil, Merck and Georgia-Pacific are members. OPAF members are the ecosystem of innovation that is designing and implementing the future of open, intelligent decision-making systems that are increasing the safety, efficiency, and agility of industry.
The issue of proximity and critical decisions is going to drive, literally, trillions of dollars of IT spending over the next decades. Billions of IoT devices will be deployed as applications and data get moved to the “Edge” – or closer to the physical world. And “closer” means exponentially more micro-clouds in stores, lamp posts, cell towers and the tops of skyscrapers. Even the mile-long oil and gas refinery requires several clouds on-site to ensure critical applications can react within milliseconds to sensor data.
The flexible and ubiquitous tools for applications for virtualization in cloud computing infrastructure are well understood and well-loved today by IT managers and the DevOps community. However, clouds typically are big, ponderous things – and often consist of tens of thousands of servers in a single location, usually wherever power, taxes and real estate are cheapest. Large clouds in one location are also relatively easy to manage, sell and consume. In fact, we call them Public Clouds because they are practically like a utility.
Solving for Proximity
Finally, we get at the heart of the solving the technical challenge of “proximity”: the orchestration and management of Edge Clouds. Instead of 30,000 servers in one location, there are three servers in 10,000 locations (micro-clouds), and you have created a massive management and orchestration challenge. How do you authenticate a user? How do you deploy an application? How do you know what resources are available? How do you connect or reboot an IoT device? How do you securely network from one application or user to another, or several others? And without an army of people… how can you still get to use your familiar dev-stack tools.
It gets really complicated really fast, but that is what a comprehensive Edge Cloud orchestration and management platform solves. And that is what the disciplines of distributed computing, distributed systems, and now software-defined networking have been working to solve for the past thirty years. It’s an exciting time to be a part of the new ecosystem of growth and innovation in Edge Clouds. Space probes are getting more automated in their decision making all the time. College students?
Read our newest white paper on multi-site OpenStack cloud orchestration to learn how we enable enterprises to quickly and reliably deploy a network of globally-distributed clouds.