Assured sUAS Autonomy Research Initiative

RCGroups user Topdogzrc sent me this and I just had to share it. It sounds like such a great project that needs a modest funding goal to be met.

Splash

Help Support this Cool Project

RCGroups user Topdogzrc sent me this and I just had to share it. It sounds like such a great project that needs a modest funding goal to be met. Check out the project below and click the support link at the bottom if you are willing and able to help.

Project Background

The Redundant Flight Computer (RFC) project focuses on enhancing the reliability and safety of small Unmanned Aircraft Systems (sUAS) by creating a redundant flight control system. The proposed system would serve as a “back-up” to the primary flight computer in the case of an in-flight loss of communications or control. The RFC project is part of a NASA-supported research initiative to enhance the safety of sUAS flying in the national airspace system, and allow the FAA to reconsider beyond visual line of site (BVLOS) sUAS operations.

A secondary goal of this project will be the development of an efficient and low cost variable-speed for propeller for sUAS integration. The use of variable pitch propellers in larger aircraft has proven to be an effective tool for increasing endurance, range and efficiency.

Why is Redundancy so Important in Aviation Applications?

Aviation related accidents and incidents often place a heavy burden on industry and can carry significant economic, social and regulatory impacts. As technology evolves, so does our understanding of how systems should be developed to minimize risk and lower the probability of failure. The Federal Aviation Administration (FAA) is currently assessing the need for full-telemetry feedback and redundant flight control systems for sUAS in commercial operations, with emphasis on aircraft conducting approved Beyond Visual Line of Sight (BVLOS) operations.

The RFC project proposes the design of an experimental sUAS systems architecture. The project introduces the integration of a secondary flight-computer that provides full telemetry feedback from integrated systems components, such as servos, electronic speed controls (ESC), autopilot hardware, etc. This system can be used to provide secondary control of the aircraft in the case primary auto-flight components fail.

What is the Redundant Flight Computer System?

The RFC System consists of an experimental (sUAS) systems architecture. The system is comprised of software and hardware components strategically placed to always maintain communications, command and control (C3) of the aircraft when the primary auto-flight system has become compromised.

The project introduces the integration of a secondary flight-computer designed to maintain control of the sUA if the primary flight computer fails. The RFC provides full telemetry feedback from integrated systems components, such as servos, electronic speed controls (ESC), autopilot hardware, etc. The FAA is currently assessing the need for full telemetry feedback and redundant flight control systems for sUAS in commercial operations with emphasis placed on fixed wing platforms conducting BVLOS operations.

What about Variable-Pitch Propellers?

The secondary goal of this project is to introduce variable-pitch propellers to sUAS applications. Larger aircraft are commonly equipped with variable-pitch propellers, which enhances performance in all phases of flight.

Why Build a Redundant Flight Control System?

This project will be highly beneficial by addressing two sUAS limitations hindering the industry. The RFC System will work to achieve a more robust systems architecture which current sUA don’t have and will enhance sUAS BVLOS operations. It will also improve flight performance by utilizing a variable-pitch propeller system, and between the two advances in systems architecture, the project will allow sUAS platforms to safely become more prolific in our airspace.

The approval of Beyond Visual Line of Sight (BVLOS) waivers, administered by the Federal Aviation Administration (FAA), is extensive. It depends on the design of a system that can self-diagnose issues and notify the operator of in-flight automation failures. Many sUAS rely on a single flight computer, which in the event of a failure, may result in the total loss of the unmanned aircraft. A redundant flight control system will ensure the aircraft returns to its launch point in the event of primary command and control issue, which may sUAS reliability and safety.

By utilizing a pixhawk-controlled variable pitch propeller system, the RFC shall solve the issues of fixed-pitch propeller systems that are most common in sUAS applications and cannot be adjusted during flight operations. The ability to adjust propeller pitch may create efficiencies that enhance climb performance, range and endurance.

Jeti Ditex Telemetry and its Advancement of sUAS

When Jeti and Hacker released the Ditex telemetry servos along with the CB210, the systems opened the door to allow an innovative sUA architecture. This architecture allows for a central power distribution system that has multiple inputs for telemetry devices such as the Jeti Mezon ESCs and Ditex servos, and it also allows for two flight computers to act as a primary and secondary device to ensure the sUA always returns home if one flight computer fails.

Objectives of the Project

  • Verify telemetry systems of the RFC
  • Verify self-diagnosing of failed components
  • Ground test the RFC hardware
  • Flight test the RFC hardware
  • Develop the autonomously controlled variable pitch propulsion system for sUAS applications
  • Build 125 hours across three testbeds to ensure reliability of the RFC

Why We Need Your Support

With your support, the RFC group will be able to fund the purchase of the components needed to build the RFC system, as well as meet the required $2,000 goal set out by NASA. Your donations will help bring to life an opportunity for the advancement of the sUAS industry and help solve concerns that affect successful BVLOS flight operations. Receiving full NASA approval to proceed with the project is a prestigious honor that will provide real-life sUAS development experience for the RFC group and result in valuable career and research opportunities. Support the RFC project and help us achieve our goal of obtaining the funding needed to receive NASA approval to proceed and better the future of the sUAS industry.

Fundraising Link - http://givingto.erau.edu/assuredautonomy

Disclaimer: This material is based upon a proposal tentatively selected by NASA for a grant award of $28K, subject to successful crowdfunding. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of NASA

Thread Tools
Aug 20, 2020, 12:51 PM
Wanted: More Time to Fly
Topdogzrc's Avatar
Thank you so much Jason for setting this up and for helping raise awareness of my research project to the RCGroups community!

I will be posting periodic updates on how things progress with the project, and we're slated to start this upcoming Monday. If anyone has any questions about how the system works and why I am creating this system, do not hesitate to reach out and ask me.
Aug 20, 2020, 01:13 PM
RCG Admin
Jason Cole's Avatar
Thread OP
Thanks for keeping us up date to date on the project. Good luck!
Aug 20, 2020, 02:34 PM
Flying Hazard
SU-4ever's Avatar
This is a great project, I'll keep an eye on it, great article!
Aug 21, 2020, 12:32 AM
Registered User
This is great!

Just one question: what is your strategy for the management (voting) in case of failure? How do you guaranty that you switch a/the failed system to a/the health one?
Aug 21, 2020, 03:09 PM
Wanted: More Time to Fly
Topdogzrc's Avatar
Hello fnev,

So the basic explanation is the primary pixhawk flight controller shall be in control until the secondary sees that it is failing or exceeds preset limits, at which point it'll trigger the primary pixhawk into a failsafe condition so the secondary can take control and return the sUA safely to the launch point.

Below explains more in depth about what my objectives are over the coming year with the tests:
Over the course of the project I'll be conducting tests using several different methods to verify the switch-off from primary to secondary pixhawk occurs successfully and stays on the secondary pixhawk, but the end test will be the one I will stay with once implemented. With this system, the end goal is to fly with the primary flight controller unless it suffers a failure, which then the system will switch to the secondary controller to allow the sUA to return to the launch point safely.
One major factor to note: I will be starting with SBus protocol for communication between the CB210 and the Pixhawks which will not have telemetry feedback like ExBus, but as it seems SRXL2 is quite similar to ExBus per tests I've seen, so I plan to implement ExBus into the Pixhawks

The first basic proof-of-concept test will be a tertiary physical kill-switch to stop communication between the CB210 and the primary Pixhawk so the CB210 will be forced to hand over control to the secondary pixhawk, and verify the switching method works as intended.

The second test will be triggering the primary pixhawk to enter a failsafe setting, and then verify it stays in failsafe while the secondary flight computer flies the sUA home. The settings to enter a failsafe shall be on the primary pixhawk and then it shall be put into a situation where it will be forced to enter failsafe.

For the end system, I will be using UAVCAN (CANBUS) to allow for communication between the two pixhawk flight computers to ensure if the primary pixhawk exceeds set limits (geofence boundaries, airspeed, pitch/roll limits) the other pixhawk will trigger the primary pixhawk into a failsafe condition.

So in essence, the secondary pixhawk will function as the overseer of the primary pixhawk to ensure if it fails, it can always take over and return the sUA home. From my understanding of the CB210 architecture too, that's how it functions, with a primary and a secondary RX input.
Aug 23, 2020, 12:05 AM
Registered User
Very interested to hear how this turns out.
Aug 23, 2020, 12:11 PM
Wanted: More Time to Fly
Topdogzrc's Avatar
Quote:
Originally Posted by rcsoar4fun
Very interested to hear how this turns out.
I'm hoping I can regain lab access this upcoming week so I can assemble the system and post some videos demonstrating the basic concepts of it. Lab access was suspended this summer so we'll see what happens, but given how all of my components are locked in that lab, I need to gain access.

And I hope many in the sUA community feel that there will be a need for this system. I personally have had pixhawks and APMs fail in flight so having a secondary one to take control would've been highly beneficial during those tests gone wrong.
Aug 23, 2020, 12:17 PM
Registered User
Have you considered some of this development in the context of the blue UAS efforts currently underway?
Aug 23, 2020, 03:02 PM
Wanted: More Time to Fly
Topdogzrc's Avatar
Quote:
Originally Posted by rcsoar4fun
Have you considered some of this development in the context of the blue UAS efforts currently underway?
I assume you meant to say BVLOS efforts?
If so, yes I have. I am a founding member of a UAV company that is spearheading BVLOS development, but I stepped down to return to finishing a degree at the end of 2018 so right now I'm an undergraduate student. I still speak with the heads of the company, help them out when needed, and I keep versed in the FAA's efforts to allow for BVLOS in the US after they did a total reversal beginning towards the end of 2018/beginning of 2019 as they had gathered enough data from the BVLOS waivers they had granted to select companies in 2017.

The three largest issues are sense/avoid of course which is where Remote ID came in, Unmanned Traffic Management (UTM) for urban environments, and system reliability/redundancy as this latest generation of sUA use components with limited reliability data/open source hardware/software which raises concerns.

As this fundraising effort required is to gauge industry/public interest, the heads of the research challenge are mainly looking at which companies/people support the project. So far I do have industry interest in this concept from a few companies, as this helps with the system reliability/redundancy concerns.

Of course, down the line I am looking to integrate it with some of the sense/avoid systems, like the Iris Casia sense and avoid system, along with Auterion Skynode which is a flight computer capable of handling the processing power of sense/avoid, and then use the skynodes in a primary/backup configuration like what I'm using with the Pixhawk 2.1s currently.
Aug 24, 2020, 07:31 PM
Registered User

best uav for kids too


best idea you have gotten on hear. start a kickstarter or gofundme.
Aug 24, 2020, 11:45 PM
Registered User
Topdogzrc, your reply to my question might seems logical at first. I am not sure if, especially in case of BVLOS operation, that the authorities controlling the airspace won't be requesting a more robust system.

Are you familiar with the failure analysis/mitigations documents that you will have to provide to these authorities?
Aug 25, 2020, 07:08 PM
Wanted: More Time to Fly
Topdogzrc's Avatar
Quote:
Originally Posted by fnev
Topdogzrc, your reply to my question might seems logical at first. I am not sure if, especially in case of BVLOS operation, that the authorities controlling the airspace won't be requesting a more robust system.

Are you familiar with the failure analysis/mitigations documents that you will have to provide to these authorities?
So interestingly enough, the FAA is looking for test data to verify systems and their reliability. The only unclear thing that I'll need to get a final answer from my FAA contact is whether an FAA rep needs to be there for all of the flight tests.

So for the first year of testing, the end goal is to have 3 airframes accumulate 125 hours total (at a minimum) to get baseline data, but then have the airframes accumulate 125 hours of flight time each for 375 hours total, and then use that data for the FAA to make a decision by submitting it to them. As systems reliability for small unmanned aircraft is still a huge question, moreso with open source hardware and software (pixhawk/arduplane), this should give them some initial answers for the system I'm developing.

Edit: to clarify, I am familiar with the documents I'll need to provide, but they're looking for test data primarily right now so they can analyze it.
Last edited by Topdogzrc; Aug 25, 2020 at 07:20 PM.
Aug 25, 2020, 07:09 PM
Wanted: More Time to Fly
Topdogzrc's Avatar
Small update as well:
Now that all of the parts are in, we plan to begin Phase 0.5 to conduct limited risk testing with a Skyhunter airframe.
Can't wait to gain lab access so we can get some videos of the system in action!
Aug 30, 2020, 04:26 PM
Purdue Engineering
Rocketman1092's Avatar
I'm unsure I understand the role of the Jeti system in this. Servo and ESC telemetry is neat, but Ardupilot already directly supports this when using UAVCAN servos and BLHeli or UAVCAN ESCs. And as far as the failsafe/failover scheme goes, using this thing more or less moves the single point of failure from one of the Pixhawks (which already has redundant sensors and EKF/DCM lanes) to the Jeti device.

Everything you've proposed seems well thought out, but I think it might be tough to prove that it actually reduces the risk of flight-critical failure. Redundancy is hard!


Quick Reply
Message:

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Build Log Model Slipways Assurance class steam tug for me srama Scale Boats 39 Sep 23, 2019 03:21 PM
Discussion sUAS Recreational income versus sUAS Commercial income roids Model Aircraft & Drone Advocacy 202 Jul 17, 2019 06:33 PM
Discussion More autonomy. Clues appreciated jferfre Multirotor Drone Talk 5 May 08, 2016 03:41 PM
Contest AUTONOMY 2016 - a new soaring contest in Italy lucianozan Electric Competition Soaring-F5J/F5K/ALES/E-Soaring 3 Jan 04, 2016 04:32 AM
Discussion The Israeli concerns are exaggerated, Vladimir Putin assured his guest. Broken Wings Life, The Universe, and Politics 8 Sep 22, 2015 07:55 AM