A new computer networking approach called edge computing is promising to radically accelerate UAS workflows. Simply defined, edge computing distributes the task of computation to the edge of the network, generally very close to the sensor. In terms of UAS, edge computing accomplishes the post-processing of sensor data right on the UAS itself.
Think about how this could impact your post-mission workflow. Instead of uploading the data to the cloud or to a local server and waiting hours (days?) for the data to be processed, the data could be processed in real time as it is collected on the UAS and, after a QA review, shipped it off directly to the client. Real-time data would be a boon to the construction industry where virtually instantaneous feedback could help accelerate a host of work processes. Operators flying inspection missions of all kinds could alert asset owners of mission-critical flaws in real-time for safer and faster inspections. These are just two examples, but seriously, where wouldn’t faster turnarounds help? I can’t think of a time when a client said “Take your time on that project, no hurry.”
Now, you may think this idea lies in the realm of the distant future, but Moore’s law and the ever-shrinking size of computer processors are bringing the possibility of onboard UAS post-processing into the near term.
Consider the introduction of the Manifold 2 on-board super computer. The Manifold 2 is available with either the Intel Core i7-8550U CPU or a NVIDIA Jetson TX2 supercomputer. According the manufacturer, Manifold 2 allows “users to process complex image data onboard the drone and get results immediately, and can program drones to fly autonomously while identifying objects and avoiding obstacles.” The NVIDIA GPU is ideal for AI, object recognition, and processing complex data while the Intel CPU has the edge on autonomous flight and robotics.
This is just one example; I suspect other UAS manufacturers are currently in the process of developing this technology and will be incorporating it into their platforms in the near future. I’ll be attending two UAS conferences this fall and looking for more news on this front; stay tuned for further developments.