Menu

Colin Devroe

Reverse Engineer. Blogger.

Attending the 2017 Pennsylvania sUAS Expo

May 11, 2017

Acronyms are all the rage these days and so it can be tough to keep them all straight. Don’t be ashamed if you have no idea what UAS stands for. I didn’t either.

UAS stands for Unmanned Aerial System. Like an Unmanned Aerial Vehicle (UAV) an UAS involves more than simply a vehicle and usually also includes camera, or multiple cameras, various sensors and other instruments, etc. Why the need for this other acronym? I’m not really sure but I believe it is to denote that these systems are generally more complicated and nuanced than your typical hobby “drone”.

The “s” in sUAS stands for “surveyors”. I was wrong, it stands for “small”. Thanks Frank.

As I’ve said in the past on my blog… I don’t mind using the generic term “drone” (even when referring to non-autonomous flight vehicles) so I will for this post too. However, I’d like to make one addition my own personal use of this word on my blog. Now when I refer to a drone I may also be referring to a fixed-winged system rather than your typical quadcopter or propeller style.

With that housekeeping out of the way, let me tell you about this expo.

The 2017 Pennsylvania sUAS Expo was in State College, PA at the Penn State campus in the Penn Stater building and was run by the Pennsylvania Society of Land Surveyors. The building was excellent for this type of conference and I really wish we had a facility like it in Scranton.

I felt like I had a pretty good handle on many of the topics, services, and uses of UAS that would be covered at this expo. However, I found that my knowledge had a ton of gaps in it and was out-of-date. So the vendors and experts at the expo ended up filling in those gaps and bringing me up-to-date with what is possible in the industry, what challenges it still has, and also how the applications for this technology is still being explored.

I still feel like we’re in the infancy of how UAS will be used. We’ve all seen Amazon’s drone delivery videos. And land surveyors are already taking full advantage of these affordable and incredibly capable systems to bring down the cost, reduce the time, and enhance the practical applications of their work.

Here are a few things I learned.

  • The algorithms for stitching together multiple photos to create an accurate 3D spacial platform are getting incredibly good. Following best practices with on-site markers, good optics, and being comprehensive in your coverage of an area, a surveyor can get accuracy levels to be within centimeters.
  • That being said, there is still much room for improvement on the cloud services front. Most of the enterprise level services offered do not yet have a cloud component (though all of them said they were coming soon). This means for every 20 gigapixels of mission photos you’ll need an entire day to process on a typical GPU. And, most of the apps do not support multi-threading so they cannot span that across multiple GPUs on the same system. You can, however, process it on multiple machines paying for a license on each. So processing time could be greatly reduced using the cloud and likely the cost also.
  • Also disappointing was that these cloud services do not take advantage of any advanced machine learning or crowd-sourced mapping to help the client-based algorithms to get better at their edge detection, transparent surfaces, or water detection. When I brought this up it didn’t even seem to be on their roadmap. Which I found odd. These apps need to improve very quickly to stay competitive.
  • The general public is still under-educated as to the deliverables for these missions. What I mean is, it seems the surveyors typically have to show their clients how incredibly useful this data can be for them. An example given was a hospital with a brand new roof had no idea they had been struck by lightning. So public education on these services has much work to be done.
  • On the surface the toolset available to most surveyors seems complex and unaffordable. Currently it can cost a brand-new business nearly $20,000 just to get up and running. I believe we’ll see this cost plummet by nearly half and the capability of the software and services to increase exponentially in the next 5 years. So I believe the real opportunity to get involved in this is over the next 24 months.
  • The people and companies to best bring this new technology to the market are the ones that already serve the customers that use surveying on a regular basis. Most often these industries have rigid requirements, data security policies, and other specific things that – if a company already has these measures in place – can be a real competitive advantage to someone just walking off the street with a UAV under their arm.
  • While the hardware and software are getting better and better it still does not allow the pilots or data crunchers to get complacent in their process. The more serious someone takes each of these steps the better the data will be and the more valuable their services will be as well.

I enjoyed attending this expo and look forward to how I’ll use what I learned to help Condron Media to service these industries and the surveyors themselves.