Hello everyone,
Over the past six months I’ve shared with you the progress of our ‘flying ball’ project Fleye. So I imagine you’ll be interested to know that we are now live on Kickstarter. The video tells our story and the page explains in more details the technical specifications.
In addition to the form factor, Fleye is also interesting on the computing performance side. We have a hybrid autopilot, using a cortex M4 for time sensitive control systems and a ARM A9 on board Linux computer to run custom drone applications. We will provide an API and SDK enabling developers to write custom application running directly on the drone.
Our platform is based on an iMX6, which has a GPU supporting OpenGL/OpenCL, and we support OpenCV, the popular computer vision library. This means that you can write applications that leverage the video feed to take actions, such as color/tag detection/tracking, face detection, etc.
Fleye is thus quite different than a classic RC controlled drone. We prefer to call it a ‘flying robot’ since it opens many possibilities to experiment safely with a fully autonomous flying platform.
What do you think? I would love to get your feedback and I remain of course available to answer your questions.
For the Fleye team,
Laurent
Comments
Well ....Sometimes it is not going according to plan :-(
''We are not in a position to deliver the product as initially promised and we have decided to cancel the project'
source: https://www.kickstarter.com/projects/gofleye/fleye-your-personal-fl...
Hi Rob,
Actually the ram drag comes directly from Newton's second law. When the air flow to enter the duct has some relative horizontal velocity (when the drone is advancing or hovering with lateral wind), the duct tends to remove that horizontal velocity as the air flow has to leave the duct along the duct axis. It results that the duct acts on the flow to stop its horizontal motion and as a counter-reaction the flow acts on the duct, this is the ram-drag or acquisition drag.
The formula to compute it is F(N)=Q.V.cos(alpha) where Q is the mass air flow (kg/m³); V the relative horizontal velocity (m/s) and alpha the angle between the duct axis and the relative velocity.
Hence, as the duct leans into the wind to go faster, the alpha angle decreases and the ram drag also.
This is only for the ram drag itself, what makes it more complicated is the associated moment which tends to pitch up the drone and impeach it to gain velocity, because the virtual point of application of the ram drag is located somewhere above the upper duct lip.
I hope it was clear ;) there are many papers on the topic, you can check the following one if you have time : §2.2.5 of http://www.engr.psu.edu/rcoe/theses/Eric_Tobias.pdf
Dimitri (CTO at Fleye)
So I did some research on ram drag, but I'm not sure how it applies here? Ram Drag is created when ducting is used to cause the velocity of the incoming air to compress the air, such as into the compressor stage of a turbine. Usually, reduced area would be used. Are you compressing the air internally for some reason? Or is there reduced area in the throat of the duct?
Design without compromise has a name - It's called Art.
Keep up the good work!
R,
C
Bonjour Laurent,
Very interesting project, and a great technology challenge, I would appreciate to get your feedback on these:
- What about using a counter rotating configuration, it has demonstrated advantages as zero counter torque, gyroscopic moment elimination and better yaw control?
- Do you have any active control at the lip boundary to compensate for the distortion the inlet flow with elevated forward flight velocity or crosswind or you will only use the vanes... and a very fast and clever algorithm...
- What are the Freescale processor offered , I am guessing the i.MX 6 Solox for the standard unit and the i.MX 6 Quad for developers ?
Freescale i.MX6 series are really good Application Processor for autopilot and I cant wait to see how you will exploit all the possibilities. I saw the presentation video showing the A9 rebooting while the M4 still running.. awesome.
Bon Succès !!
@greg You are spot on about engineering compromises. Another example is that we could remove the safety grids and have longer flight time, but that would become a finger cutter :-)
As for the slow speed, this is in fact not really due to the form drag, but another phenomenon called the ram drag effect which has an impact one order greater than the form drag. There are however ways to mitigate it and we keep working on it.
@rob The wind resistance is impacted by a phenomenon called the 'ram drag' effect, and not so much the form drag. This ram drag makes it difficult to have large tilt angles. It can however be solved by careful design, in particular on the CG placement and improving the overall control authority. In the end, we could have tilt of 90 degrees, making it an 'annular wing' and having excellent wind resistance, but then the issue becomes the placement of the camera. There is thus some margin for improvement, even on this first version, but we wanted to be honest and have communicated our current capabilities.
@coby Thank you very much for your support and nice words! In terms of efficiency (thrust/watt), the duct clearly improves things compared to a propeller alone, if the duct lip and propeller shape is optimized, this compensate for the added weight of the duct. What really penalizes us is the drag induced by the top and bottom 'finger saving' grids and their weight. We aren't at the optimum and know there is still space for improvement on that.
Looks like it tilts over 15 - 20 degrees to get to a fast walking speed. Lots of side area to catch the wind. Defiantly not built for speed, but then would you fly a quad down the office hallway past people? It's more of an indoor machine. Speed was traded for safety. Everything in aviation is a compromise.
-
1
-
2
-
3
of 3 Next