Feature: Flight School
Taking a Surveying UAS for a Spin
By Gavin Schrock, PLS
When I was asked if I’d like to attend a flight school for a UAS designed for surveying, my “inner geek” was jumping up and down—of course! Sure, there are a lot of unanswered questions that surveyors and others have about these unmanned aerial systems, like: How viable and precise are they for surveying? Would this be a good addition to surveyors’ toolkits? Why is this called a UAS and not a UAV? Are we allowed to fly them in this country? We addressed some of these questions in “Yearning to Fly
,” in the July print issue of Professional Surveyor Magazine
where we examined a commercial UAS. This article focuses on the all-important key question: How hard are these things to learn how to fly?
The June 2012 flight school was a five-day course held in Calgary, Alberta, and hosted by Cansel
, one of the largest Canadian survey/engineering/construction sales and support companies. The course was specifically designed to bring a novice up to speed on safe and effective operation of a UAS designed for surveying, specifically the Gatewing X100 (highlighted in the July article). Gatewing
, recently purchased by Trimble
, sent three trainers from their Belgian headquarters to teach sales and technical support staff from both Cansel and Trimble (and I heard there was plenty of jockeying for coveted spots in this first course). The North American market for UAS and the Gatewing X100 is now just opening up, and many are eager to get their hands on them.
Calgary was a logical place to hold one of the first flight schools in North America for this particular UAS; the Canadian aviation rules are quite a bit easier to deal with than those currently in the United States. Calgary is also a hub of activity in support of the lucrative “oil patch” of the Canadian North, and southern Alberta also has substantial agricultural lands ripe for another major use of commercial UAS. The course was held at the local Cansel office and in some nearby grazing fields, which was arranged between a local farmer and Cansel.
One thing that became rapidly apparent in the course of the flight school was that students had to be prepared to discard a lot of preconceptions, primarily that they would be “flying” the craft. In reality, as we soon learned, the role of the operator is to follow meticulously the procedures prescribed to send this little UAS on a mission, bring it to a safe return, download the photos and log files it generated on the flight, and post-process the same into surprisingly precise and valuable surveying data.
Experience Not Necessary, but It Helps
You do not need any specific skills or experience to operate the UAS, though you will need to become versed in the respective civil aviation rules and regulations for where you wish to operate it (more on that later). If you can follow procedures carefully and faithfully, then almost anyone can operate one. The procedures are everything—there are well-defined checklists that must be completed before leaving the shop and once more in the field, pre-flight. These procedures ensure that the system is fully operational and in top condition, that the flight plan is properly thought-out and pre-programmed, that local field conditions are taken into account, and that the operator and spotter are ready for action.
The five training days were a mix of two initial days of classroom fundamentals, two more days of field/flights, and another classroom day of processing. The first two days in the classroom included familiarity with the X100, how it operates, key features, flight pre-planning, camera settings, a few aeronautics basics, an overview of applicable civil aviation rules, download and processing procedures, and the very important care and handling of the craft. Yes, we were all chomping at the bit to get out there and see what the X100 could do, but we were very glad for the classroom time when we actually hit the field. The UAS is essentially a camera system that flies; to get the most out of the unit and to reap optimal cost-benefit the team at Gatewing have put years of testing and production flight experience into the procedures and checklists.
Some of the fundamentals that were taught included a pre-flight test of the accelerometer/potentiometers that inform the ailerons. Gatewing support engineer Alain Quackelbeen demonstrated the checklist step of manually simulating pitch and roll while observing the aileron and servo-motor reactions. Checking the air pressure reactivity of the Pitot tube (essential for gauging airspeed) was another step. We all asked the obvious: Would some actual flying experience be helpful in grasping fundamentals like these? Of course, we were told, but the explanations and importance of how to incorporate aeronautical fundamentals into the flight planning were sufficiently well-covered—even for novices without any prior flight experience. But, like any new tool that surveyors bring into their toolkits, it does pay to be a bit of a geek.
Likewise, there’s the assumption that having some experience piloting radio-controlled (RC) model aircraft might be helpful. Perhaps it might be helpful, but apart from emergency, minor in-flight adjustments to avoid other craft or obstructions in the flight path and landing pattern (and even these are limited to some tweaks or “bumps” executed from the control tablet), you do not actually control the UAS in-flight. The X100 is flying at a relatively high speed, typically 80km/hr or more for stability and to track along well-defined flight patterns to ensure good overlap of photos, adjusting for cross winds as it goes. A key message of the training was to let the UAS do its job, have a spotter track it, and be prepared to intervene, but only if need be.
Could someone directly pilot one of these effectively? These units were not designed for direct control flight. Nearly every element of design is geared toward the high-speed, highly stable flight patterns needed to accommodate the photo runs. But as an interesting aside, our Belgian trainers from Gatewing told us of one RC model champion who goes by the competition moniker of “Doc Insane”—none other than one of their own designers from Gatewing’s engineering group, Maarten De Moor, who has excelled in manually piloting a modified X100. If you’ve ever done any RC flying, you have to see one of his videos (www.youtube.com/watch?v=BuMUn-6LagY
). Don’t try this at home; unless you are a champion RC pilot, it would be best to let the X100’s autopilot do the job.
Some knowledge of the camera system would come in handy to get the most out of UAS flights. Though camera fundamentals were well examined in the classroom by Gatewing support engineer Maarten Durie, there could be a lot gained by having some additional knowledge of the subtleties of photographic elements like aperture, shutter speed, and ISO; auto settings are not recommended and would introduce too much potential latency in the shots with the craft moving at such high speeds. Camera focus and settings are made prior to the flight, informed by the proposed height above the ground and local light conditions.
Making Each Flight Count
Though the kit comes with two batteries, each capable of powering a full 40 to 45 minute flight (depending on local conditions like wind speeds), you want to plan each flight carefully and avoid long waits for recharging. There is also an indirect cost per flight in that the replaceable fuselage lasts around 50 flights, depending on how well you can plan and execute landings and on what types of surfaces you land on. Even though the cost of a replacement fuselage is only about 10% of the total cost of the UAS, you will want to make each flight count. There is, of course, fear that you might lose a UAS that costs about the same as a total station, though with the standard locator beacon included, the Gatewing team have reported that there have been very few craft completely lost in years of flights—pre-flight planning is the key in getting your bird home safely.
Even before you head out to the field, you can examine potential flight plans by downloading an aerial or satellite image of the area to be surveyed, even adding Shape files (.shp) and KML files to lay out in the provided mission software where the “blocks” are that will represent individual flights. Each of these “blocks” would be around 1.5 km² at a height above ground of 150m; this size would vary depending on desired height. Flight planning is done on the Trimble Tablet and Quickfield Flight software included in the kit. You can lay out “block” options over the aerial/satellite image, account for different wind directions and obstacles, vary the desired height, and the resultant expected pixel size and precisions are displayed.
We realized once in the field that a check of the local landing surface and day-of-wind direction helps refine the flight plan, which can be adjusted easily in the field in the Quick Flight software. We even had a few instances of last-minute wind-direction changes to the plan. It is preferable to have the UAS take off and land into the wind to provide optimal stability for these two critical phases. The flight path during the photo taking is across the wind; this avoids a compression or extension of time between photos if the photo legs were to have been upwind or downwind. We did have one flight where the wind direction changed by roughly 60 degrees during the course of the flight. Even with the crosswind (and light rain) the X100 landed about three feet from the proposed, GPS-logged landing spot (ironically the best landing of the day).
Assembly, Pre-flight Checks, and Launch
Once finally in the field, having marked desired take-off and landing spots with the GPS in the tablet, visually ensuring the prescribed glide path and approach were clear of obstructions, clearing the area of looky-loos and cows (we had to shoo some away, though the cowpies remained a constant hazard), we began the final pre-flight checklists and assembly. To make sure that key components are working properly, they are installed into the fuselage on-site. The brain-box, containing the autopilot, GPS, accelerometers, pressure sensors, and the radio (802.11 standard 2.4GHz) are secured. The locator beacon is inserted, and the companion detector handheld is tested. The camera is adjusted and inserted into the payload bay (there is an optional infrared version as well)—pretty much each and every step of the checklist for the shop is repeated in the field.
An additional step in our field training was the setting of aerial targets, recording the positions with a rover receiving corrections from the Can-Net VRS Network (operated by Cansel). These positions enable precise registration of the orthomosaic of photos recorded by the flight. Resolution of the orthophotos depends on the height above the ground flown. At 150m the pixel is about 5cm, or a little over 3cm at 100m. Planimetric accuracy principles apply as would any aerial photography; the mission and post-processing software provide statistical summaries based on flight-specific parameters. While the addition of ground targets does not necessarily alter the relative accuracy, it does provide the georegistration that the navigation-grade GPS does not.
The catapult included in the kit is then assembled. Strong elastic cords are stretched by a hand crank. The accelerometers in the UAS detect the snap of the launch and start the motor the instant the X100 leaves the end of the catapult. The motor is controlled by an electronic pulsing unit that delivers precise power adjustments and also ensures that the propeller will not slash your fingers. A rear drive engine is the best design for catapult launches, and the folding propeller blades push the wing-load-designed craft so it flies more like a jet fighter than a glider with higher speed and stability in mind.
The observer visually follows the craft in flight, and the operator monitors the progress of the flight along the pre-defined photo-run legs on the tablet, at the ready to make any in-flight adjustments or “bumps” (arrow keys on the tablet interface) to avoid hazards like passing aircraft, trees and buildings, or birds. Teams in the course took turns making such adjustments in reaction to simulated hazards and tried the mission abort options; these bring the craft immediately to a safe landing pattern. If communications are lost, the X100 will automatically enter into a landing pattern.
We had the pleasure of trying flights on a bright and sunny first day of light winds only to have a damp and windy second day of flights. The experiences of the miserable second day certainly answered questions about operating in rain and varied cross winds. If the winds are under about 30 knots and the rain is not pelting, then the X100 performs as splendidly as it did on the second day. The finale of the field training was for each team to execute a complete flight from planning through a successful landing without the trainers intervening—we were nervous but successful.
Bringing It Home
Barring unforeseen circumstances or poor planning, the X100 will bring itself back to the designated landing spot as it did during our training flights quite precisely. Next is downloading the photos and flight logs. The GPS positions logged with the photos provide a rough geo-registration for the Stretchout orthomosaicing software provided; it takes over and uses pixel pattern recognition to precisely align the photos into a seamless image. Further processing in the provided software (and suggested software for more advanced needs) produces a number of formats and representations of the digital surface model cloud in text, KML tiles, and more.
And when the bird comes home, you need to just about completely disassemble it, but for good reason; the reassembly before the next flight is the proven method for ensuring all parts are accounted for and in top operating condition—get used to the checklists!
But when will you be able to bring one of these units home to your company or workgroup? You can do that today, but there are a few more bureaucratic hurdles to clear before the FAA solidifies the rules for widespread use. This could just be a matter of applying for a waiver for certain public sector entities and as little as a year from now for the other commercial uses. How will these lightweight UAS for surveying impact aerial photographic markets? Opinions solicited from aerial photography practitioners have ranged from viewing these UAS as toys to be ignored to others fearing an erosion of market share. Even the UAS manufacturers do not view such technology as replacement for either high-precision aerial photogrammetry or terrestrial survey, but it is almost inevitable that such innovations will open a new, surprisingly affordable and functional niche.
It was a great school! And apart from some miserable weather, nosy cows, and Canadians talking me into trying a plate of something called “poutine” (look it up; it is a little scary), I feel that the system and instruction are solid even for a novice. While not a direct endorsement of this specific UAS, as there are a variety of such craft available, I can honestly say that I was duly impressed with the completeness of this technology. And I can definitely see that UAS for surveying is an affordable, reliable, and fully functional, potential addition to the standard toolkits of many in our profession. When can I have one, boss?
View the related video here
Gavin Schrock, PLS is a surveyor, technology writer, and operator of an RTN. He’s also on our editorial board.
Are we allowed to fly these in this country?
Aviation rules for lightweight unmanned aircraft systems (UAS) vary greatly across the globe. In some countries it is a relative free-for-all, while other countries require a pilot’s license. The rules in the United States fall somewhere in between. These rules are about to undergo rapid changes driven by an awareness of UAS uses by the military, the viability of new commercial UAS, and some key legislative initiatives. As part of the 2012 FAA reauthorization bill signed into law this past February, the Federal Aviation Administration (FAA) is required to analyze, test, and develop guidelines for commercial UAS uses. While there are legitimate concerns by many about safety of the skies and privacy, the skies are currently full of craft that can and do pack cameras, and RC models fly with few restrictions. Not to minimize these concerns, but it is almost assured that we will see widespread commercial use of these units within a few short years.
Ben Gielow, government relations manager and general counsel for the Association for Unmanned Vehicle Systems International (AUVSI, www.auvsi.org
), summarized the current rules in the United States as follows: the FAA currently prohibits all commercial use of UAS, regardless of size; however, public entities are allowed to operate UAS by applying for a waiver from the FAA, which is known as a certificate of authorization (COA). Although a COA can be granted for any sized UAS, the FAA is easing the requirements for UAS that weigh less than 25 lbs and are used by law enforcement. Some public uses include federal, state, and local government, publicly owned utilities, and public academic institutions. For the most part, operations are limited to line-of-sight (from the viewpoint of the operator) and are restricted to daylight hours. Flights must be conducted by daylight operations and be less than 400 feet from the ground. The FAA considers these COAs on a case-by-case basis.
The FAA is setting up six test sites to evaluate guidelines for broader use of commercial UAS, and there are dozens of states in heated competition to host such sites. The sites should be selected by the end of 2012 with the first sites becoming operational in early 2013. By August of 2014, the FAA is expected to establish rules for the commercial use of small UAS, weighing 55 lbs or less. This will be the first time a private surveyor will be able to use a small UAS in furtherance of their business. A free-for-all is not expected, but what is expected is a well-defined process for public and private users to seek waivers and file flight plans with approval for uses like surveying, mapping, precision agriculture, and more.
» Back to our Aerial Mapping Fall '12 Issue