One of the bigger challenges associated with deploying Internet of Things (IoT) applications is that by definition the devices themselves don’t have much of footprint. As a result, developers of those applications are likely to turn to containers as a method to deploy lightweight amounts of code using a format that can support any number of device types and processors easily.
With that goal in mind Resion.io has raised $9 million in additional funding from DFJ, GE Ventures and Ericsson to drive development of a wireless networking framework using containers for deploying and updating IoT applications.
Resin.io CEO Alexandros Marinos says most IoT devices will be deployed on devices that are not easily accessible. As such, organizations will need to find a way to deliver patches and updates that add new functionality without having to physically touch each device. To prove that point Resin.io, at the recent Dockercon 2016 conference over a wireless network, added new thermal sensing camera functionality to a drone in flight.
Built primarily on top of Linux containers, Docker, Git and a Yocto distribution of Linux optimized for embedded systems, Resin.io allows IT organizations to employ a modern DevOps framework for deploying code across thousands of IoT devices, Marinos says.
While the hype surrounding all things IoT is running high, most organizations don’t yet appreciate the challenges associated with managing an IoT environment at scale. An IoT application is essentially a massive instances of a distributed application, with application code spanning thousands of remote devices. Because of the limited amount connectivity available to those devices, an IoT application typically will need to process some amount of data locally before passing aggregated amounts of data back to a big data analytics application running in a data center. Once that analytics is performed, it may be determined that some function being performed by that device needs to change, resulting in a need to update the code running on that device. Trying to accomplish that across hundreds, sometimes even thousands, of IoT devices represents an exercise in change management that most organizations are ill-prepared to meet.
Given the scale of IoT deployments it’s entirely probable that containers will wind up running in these environments more than anywhere else. Beyond the fact that they mostly are greenfield applications, the operations teams that typically deploy embedded systems have a long history of working around the memory limitations of embedded processors. As such, they are more inclined than traditional IT organizations at the moment to appreciate the benefits of a lighterweight approach to virtualization.
It may take a while for most organizations to bridge the cultural divide that currently exists between operations teams and internal IT organizations. The good news is business leaders are gearing up to force the issue to drive trillions of dollars of economic opportunity. In the meantime, the one thing that just about everybody should be able to agree on is the strategic role containers can play in turning IoT theory into actual reality.