BVLOS SUI Endurance Remotely Piloted Survey Drones Powered By Indropilot In Brazil

🚧Note: This article is in progress, some minor future edits might be added, as well as some extra media where is needed.🚧

Img1: The SUI Endurance Drone

Abstract

The Project consisted of four drones (SUI Endurance by Hitech), Powered by Indropilot platform to carry out multiple survey operations in Brazil, and also to be piloted remotely from Canada if needed over the internet (5G/LTE), the drones can also run autonomously to carry out these survey operations.

Img2/3: The airframe of the drone without any major modifications or added components

Project Requirements

• Four Drones that will be used for site (port/mining) survey in Brazil
• The drones can be operated locally (by RC link) or remotely (by internet 5G/LTE)
• The Drones are equipped with the needed components for that missions like gimbal, Camera, etc.
• The Drones are to be shipped before the year ends (year of 2022)

Hardware And Software

The project main components are as follows:

Software

The software used to operate and fly these drones is primarily Indropilot platform which is built in house, plus other necessary software for ground station, video viewing, etc.

The software integration was straightforward as the platform was designed to be compatible with any pixhawk based drones, although the drones were running a different firmware for the autopilot -it was Ardupilot and not PX4-, the software worked as expected as it does interface with the protocol MAVLINK regardless, so the process was simple.

Hardware

The drones airframes are provided from the US company HITEC¹ and the drone model is called SUI Endurance. The airframe provides around 80% of the functionality needed for the project, and the rest of the hardware like the cellular modem, cameras, SBC, and others were added to satisfy the requirements.

Img4: The drone hardware and adding the SBC (Jetson Nano)

The extra hardware components were integrated into the drone, after the proper calculations of the total gross weight and placement

Img5: The drones hardware assembly and integration

Testing Stage

The testing stage was separated to two sub-stages, one was to test the platform functionality on a completely new drone, the other specifically for the project requirements:

Img6/7: Testing the SUI drone, you can see me staring at the drone while she stares back at me!

Testing the software platform on a new drone

The platfrom, Indropilot, was designed to be compatible with all Pixhawk drones, including this SUI Endurance drone. The integration part was as smooth as loading the SBC's SD Card with the platform and it was immediately connected. However, there are other aspects needed to be tested before making sure the setup is complete, and those are primarily as you have guessed, the configurations and tunning, plus any issues with the initial hardware components integrations as this is the first hardware iteration.

vid1: Several videos of a successful launch/landing of the drone using the software platform (Indropilot), these flights were over 5G

Testing the project requirements

After the pervious stage of testing passed, the second one was about the requirements, and those were summarized as follows:

• Four drones hardware: The hardware second iteration imporved the first one in several ways, primarily placing the 5G/LTE antennas below (as seen below) for better cellular performance since the drone will be flying higher that the towers, more secured to the drone chasses instead of the cover, and lastly to prevent an antenna hitting the prop if it happens to bend over mid-flight. Other hardware improvements were internal, like weather insulations, better wire management, and easy access to several ports like the SIM card and the IO.

Img8/9: The first picture was the first iteration, while the second show the four drones with second iteration, and the antennas are point downward

• Testing the RC control and connectivity, that includes all the configurations required for the radio controller and the RX/TX link.

Img10: The RC controllers after configured and tested for each drone

• Testing the autonomous mode of a fully autonomous mission including the auto photo captures to be used in photogrammetry and other post-processing survey applications.

vid2: Several videos of a successful launch/landing of the drone fulfilling the project requirment, these flights were over 5G/LTE

Risks And Challenges

Like any project, especially the ones that are rushed like this one, you should expect a lot of challenges and hair-pulling moments!

First challenge

The first challenge, was the timeframe, the project was awarded, timeline decided, and the requirements been approved to have the project done within less than two months, one of which was December with all the holidays delays you would expect for any logistics needed. At some point, I needed some RC components (Four FrSky X-Lite Pro remotes plus the TX/RX TBS Nano Crossfire), and it was a big challenge to find suppliers that will deliver it within the timeframe and before the holidays. Additionally, I needed some special wires for the antennas (SMA) and the flight controller, and same issue it was a struggle to secure it. On top of all that, the client had a hard deadline that all four drones should be shipped before the the year ends!

Img11: The four drones in the testing bay

Second challenge

Second challenge was in the planning phase, the client had all the requirements and agreements with the sales team, who assumed -and as you might know, an assumption in the planning phase will turn to a risk if it turned to be false!- several points and proceeded to agree in a rush to secure the client. Well, the good news, the client is secured, the bad news, there was a lot of non-sestmatic work (aka, rush around like a headless chicken and grind for extra hours) to get the job done.

Third challenge

Third one was technical, some of the four drones were having a strange anomalies whenever the mission start to fly them. The anomalies were the thrust isn't enough to lift the drone, either one or more of the four motors. The issue was puzzling that even the drone airframe manufacturer couldn't solve it, all troubleshooting returned good, so no power, wires, motors issues were found. And even if it had been in one of the motors, it would be impossible to secure a replacement with the very short timeframe the project had. The first time the issue happened, it almost crashed one of the drones as it can tip it over while all props are still working, as you can see below I was rushing to turn it off. Since then I've been only testing with props off AND with an RC link as quick way to take back the control if anything goes south, figuratively and literally. To solve the issue however, it took me around a whole week of sleepless nights to pinpoint the issue and how to resolve it. It turned out to be that the firmware of the autopilot (Ardulpilot) had some issues despite it being the latest version that time. I patched it, changed some of the configurations, and the drones were all ready!

Another technical challenge I had in these drones were the GPS unit, it was U-Blox M8 series², despite being upgraded from M7, it was glitchy at times and as a result of that the drone was grounded and could not be tested.

vid3: The drone thrust issue that caused several delays and extra troublshooting

Fourth challenge

Fourth one was environmental, as that time, it was freezing cold as you would expect in Canada. And that was a challenge for both the pilot as well as the hardware since LiPo battery life get reduced substantially in the extreme cold, as part of that energy is used to heat up before it can reach the operation temperature.

Img12/13: The tests were done in the extreme cold condition sometimes

Special Thanks

Big kudos goes to my co-worker, Liam, a co-op student that worked with me in this project and he basically took care of the hardware integrations!

Img14: My Coworker, Liam, who worked with me in this project

Conclusion

Img15: The drones are ready to be shipped!

The project was a big success despite all the challenges, the operation team did travel to Brazil and trained the local crew there, additionally, where available, pilots from Canada can take care of the mission remotely and fly and survey the site thanks the platform Indropilot!

Img16: The SUI drone flying in brazil by the operation crew

Back to top