In every industry, operations are taking an increasingly “hands-off” approach in which operations are not autonomous (human operators are kept absolutely in the loop), but are instead entirely orchestrated from a distance (human operators may never set foot on-site). The possibility of remote operators staying informed enough to manage vast and dangerous processes is entirely a result of turning on-site observations into data that can be shipped over the web. The capability to execute operator-specified commands via powerful and precise actuators is again a result of specifying those commands as data that can be shipped over the web.
This data-ization trend will only accelerate as the multitude of lightweight, inexpensive, and yet extremely capable data collection and computation devices collectively known as the Internet-of-Things (IoT) become omnipresent in industrial settings. These IoT devices will be both observers and actuators: network connected sensors create data that can be rapidly exported to a cloud environment, and elaborate actions can be carried out by smart-controls and multi-purpose robots.
The inputs and outputs from human operators will eventually consist entirely of data, but what about in the middle? How are the cloud data repositories turned into high-impact analyses? And how are complex control scenarios planned before being implemented? Most importantly, how is all of this information effectively communicated among team-members, none of whom have been physically present at the site in question?
It is this vast terrain of perception, automated analysis, presentation, precision, accuracy, communication, and planning that happens in the middle, between observation data input an command data output that we will call home.
Specifically, in 2018:
We will use photographs of an industrial facility to both algorithmically identify corrosion problems, and build a 3D model of the facility algorithmically via the process of photogrammetry. Identified problems will be placed within the 3D model, and viewed in a cloud-based application that clearly shows the location of a problem within the structural context of the whole facility. This contextualized view of a problem will be shareable via a copy-pasteable URL that can be sent to any team-member, immediately and effectively communicating the problem in a simple visual. This will enable analysis of the problem regardless of team-member location or role, since everyone can now plainly see the problem just as if they were actually on-site. Furthermore, this data model is shows the as-built structure of what actually exists (as opposed to what is specified by blueprints), and can be repeatedly updated to show an auditable record of fixes and changes made to remedy a problem.