Daisy Kite Airborne Wind Turbine

Ram air kites spin a parachute and transfer torque to ground

Similar projects worth following
We can harvest strong and expansive wind safely and efficiently. A ring of kites balances dynamic forces with tensile structure.The kite rotary motion drives a ground based generator.For a video description

This project has been continued in a further hackaday project
and kinda superceded by new models which are now outputting over 1kW

These are design and manufacture notes for Daisy3. A flying wind turbine.

After a few tweaks and tying it to a falco electric bike wheel... it looks like this and makes power...

A few more tweaks later and it looks like this.

Here with the stack of 3 rings it is generating 100w constantly in 11mph winds. Notice how the bottom ring is struggling to hold form. It is getting crushed by torque in the bridling. I have revised the bridling to solve this for a more dynamic form and will test it soon.

Old design was a bit baggy... but I may tie it stacked on the back above the new one....

New Daisy has struts ahead of the leading edge to allow the kites further twist into wind and hold them steady in low speed. Much more reliable than the old one.

The principals are straight forward. Wind energy is sustainable. Wind is mostly high up.

Summarising the science of Airborne Wind Energy (AWE) it says:

Higher wind speed = way more power also more height = more power available to harvest.

Towers can't avoid cubic mass scaling law; more height = way more weight.

Therefore the future is flying wind turbines.

Design choices:

computer automated vs simple mechanics (I use simple mechanics but with monitoring and control options)

bouyant vs powered lift vs kite lift (So far I use kite lift but will also want to investigate hydrogen kiteoons, a steerable kite balloon generating it's own h2 aloft as buoyancy needs dictate)

airborne generation vs ground generation (ground = safer ; less mass aloft)

pumping reciprocal surfaces vs crosswind track vs crosswind rotary (rotary means continuous generation and less wear on components... needs good bearings)

Google / Makani are building a robo helicopter kite. Sounds cool. But breaks some crucial rules of safe aviation. Most other companies use ground based generators ... I intend to use a small turret. (Even though Daisy can already fly without any pole e.g. in the image above or the youtube version.)

The standard FAA / CAA guidance is to have as little fast moving hard heavy stuff in the air as possible. So I try to use only rag and string soft kite architectures. Having said that, I do fly a cut down rollerblade wheel because good energy conversion often relies on a good set of bearings. The advantage to thinking about using soft rope and sheet materials is strength to weight in tension is much greater than standard compression devices. (My designs are often like fishing or climbing equipment)

Further FAA / CAA guidance suggests you should be visually conspicuous ... It's a very bright kite to help pilots. Yet just conspicuous enough to radar. I don't want to jam nor be invisible to Radar, So hoisting a small transponder on the stabilised lift kite will help.

Since we have steady 60m AGL altitude we can improve communications. The hills in this beautiful Island wilderness ruin my phone data. Luckily with my phone set to wifi hotspot and hoisted in a kite pocket, I can log in to internet (and all the goodness that brings) from ground level. Condition monitoring and many other uses come from having developed a Stabilised lightweight airborne lifting platform.

Sharing and openness is key to this project. That's how it has got this far.

Here are some more build notes

  • 1 × bike wheel MTB hub and brake... moving onto a pancake generator hub soon
  • 1 × Lifting kite You may want to consider Mothra designs see KPower kites in Texas for that. I've used a powersled joker 36. Configured in a variety of ways to lift the rotary set
  • 3 × Driving Kites Ram Air kites sewn onto a drogue ring
  • 1 × Drogue parachute You may be able to repurpouse an old one... I made mine
  • 1 × Rollerblade wheel the bearing set is strong enough to lift the back of the drogue up

View all 10 components

  • Stacking Rings

    Roddy "Rags" Read09/29/2015 at 12:28 0 comments

    I mentioned that I intended to have a go at flying a stack of 2 rings together. Using the old baggy slower ring on top of the faster newer one.

    It works a treat. Launches easily, Inflates well, Flies steadily. I had hoped to be sharing this in more detail with the cohort of the TU Delft AWE course... but the online component of that course has now unfortunately been closed.

    This now really opens the question of scaling viability. And how to get universities and corporations to engage with that.

    Please can somebody manage me / this / Daisy / Windswept and Interesting to push the project forward?

  • Dark side of net publication

    Roddy "Rags" Read04/17/2015 at 14:49 0 comments

    Similar to Hackaday is another How 2 site instructables... Hope our good hosts don't mind me mentioning here that further instructions are availabl on

    I'll likewise link back to these instructions

  • Made My First Kite Electricity

    Roddy "Rags" Read02/27/2015 at 15:02 0 comments

    Just some back in from testing.

    I sewed a new Daisy without the back end parachute kite and ran it with a torque ladder through my falco e-bike wheel.

    It made Power. OK only a wee bit so far. I have a new challenge of making enough for my mum to enjoy a cup of kite powered tea.

    I May mix the two spinning kites together now as the back end one had advantages.

    Here's the vid

  • Communications

    Roddy "Rags" Read08/18/2014 at 14:17 0 comments

    Some of you may be wondering where the electronics are... Fear not, electronics Bsc ENG is my background so... For communications I plan to re-purpose a few old mobile phones. By installing cyanogenmod Android I can easily use the phone as a tether. so what? If my phone is inside the lifting kite. It gets a much better mobile data signal than I do on the ground... That's why I want to log into a higher altitude phone signal. 

    At altitude the phone can also be reprogrammed to network with similar airborne phones... There are other projects on Hackaday making satellite based networks. Satellkite, kitellite, call it what you will higher communications devices work more efficiently. Tethered on a kite there are loads of ways to power them including airborne generation see kitesat

    Hopefully my old phones can also simplify posting NOTAM operations data.

    RPM and performance monitoring is also easy with old phones... However, check out FALCO EMOTORS

    Being oh so cool you can wirelessly adjust the performance characteristics of the generator... yay.

    They also have a setting where the resistance of the generator can be controlled by a heart rate monitor... That's easily adapted to an rpm counter.

    Fun Fun Fun...

  • Back from the mainland

    Roddy "Rags" Read08/18/2014 at 13:53 0 comments

    Thank goodness, I'm back to the windy Isle of Lewis. I attended Leominster kite festival with Daisy... luckily I also brought a poly bag, some tiny carbon rods, some thread and tape to make a lightwind sled kite for the kids...

     It was really calm by my standards... But I learned that my Daisy architecture is still overly heavy. I can cut down on the brutal over design.. I can also look back and wonder how on earth that home sewing machine did what it did... loads of WD40.

    The kite travls really well folded into a rubble sack. An old kids umbrella cover and a bit of bungee make a good cover for the wheel... This helps to stop lines snagging a lot.

  • single line mode test

    Roddy "Rags" Read07/13/2014 at 20:03 0 comments

    The Disy ring was tested directly inline with the lifter kite.

    Configuring this way I had no direct control over the tension in the rotation axis... (actually I had a light back line behind which could take the back end down if needed)

    Anyway, Here's the "proto promo" / test vid... I just happen to have been brought up and live on a gorgeous island. So handy for backgrounds

  • Valley Testing 5/July/2014

    Roddy "Rags" Read07/06/2014 at 20:31 0 comments

    First completely tower free test run today. Superb test and loads of video feedback available on youtube.

    Ran it for around 1.5 hours. Had to reset the handle position at one point to avoid crashing.

     I had the back lines a bit tight. Must not tie thin lines directly to large rocks... I knew that. must practice.

View all 7 project logs

  • 1
    Step 1

    Choose your design after reading as much background as you can. Prepare for ridicule from everyone you know. Have a load of loft space set aside. Avoid some of my mistakes. Don't try to fly the big ring with anything other than a rigid ring (or a swivel (if you don't want power)) on the ground. Don't make the ring out of a toy inflatable trampoline. (fun and easy but breaks easily.) Don't custom make a beautiful ply ring from local timber. Don't use a bendy pole for auto alignment... It bends not follows, use a universal joint when you (and I) eventually get there.

    Keep going and get ready to have a lot of fun

  • 2
    Step 2

    Respect the hard work and dedication to a better existence given by the open source community. 

    Thanks must go to Grasshopper 3d algorithmic modeling language for rhino

    And Thanks now in no particular order to...

    David Aberdeen use of your surfplan clark-y model for Dasiy and other systems modeling

    AWE yahoo forum members. Amazing feedback and analytics

    SY fishermans coop having rope by the Kg

    Gordon Diesels rapid development of hardware

    Ivor Murrays Engineering Hardware

    Leominster Kite Festival Teaching me about light wind

    Rosie and Netty Sewing lessons

    Murdo MacLeans having coats duet nr.30 thread by the spool

    Wife, family, neighbors, kite testers, passers by for scones, scrutiny & kinship

    Point North fabrics rip stop

    Rhino3D developers


    Gmail it's not the only email system available

    Dyneema It's not the only rope available

    HQ for the symphony beach 1.3 designs

    Teachers in sports, physics, maths, engineering…

    Tea, coffee, chocolate for yummyness

    Ubuntu, Android and windows for thinky computer stuff

    Make, Letting me know other nutters were keen enough

     Arduino, raspberry-pi, os hardware makers, Light Blue Bean

    value accounting can be done open source

    Cyanogenmod for letting me tether my kite mobile signal to ground

    OpenFoam CFD I'm hoping to use it for physics modelling at home soon

    Falco emotors Rakesh is a straight talking genius. His wireless controlled hub generator system may give Daisy a near off the shelf hackable connected generation solution in itself. Skull to Rakesh.

    HTC and Samsung for android hardware. My old phones can now be used as an airborne tacho at the top end of a spinning kite... and as an internet enabled wifi hotspot at the same time. yay

    FAA for sensible consideration of regulations governing integration of airborne wind energy systems into civil airspace... no easy task

    Mother in Law for the old sewing machine

  • 3
    Step 3

    Read over this system design note...

    Daisy Project System Design


    The Daisy project is an attempt to realize competitive open source airborne wind energy generation.

    Project Executive Summary

    The Daisy project is part of the rapid AWE development program by Roderick Read, Windswept and Interesting Limited.

    AWE can power world energy demand many tens of times over.This project is open source hardware.

    Shared open design led to obvious, organic looking devices. Some biomimetics and derivatives of natural pumping organisms, formed from engineering, environmental and business needs.

    New materials have been mixed with ancient technologies to change the scale of wind power.

    The designs are massively scalable. AWE can change our relationship with energy sourcing. Everybody is invited to participate in open AWE forums, open development, open prototype testing and product development.

    Roderick Read was brought up sailing, windsurfing, surfing and kiting in a tight community on the windy, religious and ancient Isle of Lewis.

    System Overview

    • 1.A stable lifting kite, able to weathercock, optimized for smooth relative lift
    • 2.A Ram air driven kite ring system spins transmitting torque through tethering
    • 3.Ground anchoring appropriate to lift, weather cocking, and torque power conversion
    • 4.Electrical generation set
    • 5.Electronic control set

    Design Constraints

    Where possible use tension systems and constrained mass of air for structural rigidity aloft. This reduces weight requirements considerably and allows scaling.

    Optimise for continual autonomous operation, ease of handling, price to performance, power to weight, modularity, marketability, safety, minimal environmental impact of power extraction.

    Stay within FAA guidelines for AWES. Adhere to advice of TACO1.0

    Future Contingencies

    Proprietary certified radar transceivers and manual NOTAMS may provide additional cost and time restraints on installations intended to operate over 60m AGL. This would be factored into scaled system costing.

    Automated layer interfacing and panel detach and reattach systems could dramatically improve deployment of multi layered ring sets and lift kite surfaces. Proposals have been openly discussed ofthe desire for machine developments to provide for the problem outlined in these scenarios thus prior art is established in the field allowing open development unhindered by patents.

    Membrane structural analysis and certification requirements do not as yet exist for this system. FEA and CFD Models of system strengths and weak points will help to optimise and suggest where reinforcements may be needed to meet standards.

    Document Organization

    For this project will be done on Hackaday


    Roderick Read, 15a Aiginish, Isle of Lewis, HS2 0PB. MOB 07899057227

    Project References

    Springer Airborne Wind Energy book

    Yahoo airborne wind energy forum


    Hardware Architecture

    • 1.Lift
    • 2.Wind to Rotary
    • 3.Rotary to Electrical
    • 4.Tethering, Anchoring and Weathercocking


    Open developments in “Mothra” single paneled skin, arched loadpath lift kites by Dave Santos of Kitelab and Roderick Read of Windswept and Interesting allow for very large lift structures and arrays to be implemented. Mothra’s use multiple ground anchoring points for safety. Mothra’s can be anchored on radial tracks for weather alignment. Mothra power control is easily adjusted by foot anchor trimming and nose tether. A simple version will be used for initial intended market scale device prototyping and tests. Mothra kites can be flown with or without loading. Thus Daisy rings can be hoisted into place when ready to be set. A Mothra can have various configurations of lift application points, thus Daisy rings can be arrayed to effectively fill the frontal area of the arch.

    A narrower post and boom centric version is proposed for initial weather cocking lift models. Wide systems as shown will be applicable to small human movable beach installations.

    Wind to Rotary

    Very large ram air and inflated kites have already proven successful up to and beyond 300sqm approx 2MW with simple kite window position flight optimization of a single line kite from a ship. The benefits of cross wind kite flight dynamics for stationary ground generation devices are impossible to ignore. Rotary systems which keep the kite in the downwind power zone are ideal. Tether drag is a serious problem for fast kite energy conversion. A rotary system does well by avoiding tethering at the outer tips of rotation.

    To maintain all of the most desirable flying and energy capture dynamics a ring system of kites was devised.We simplify the design to a parametric algorithmic describing the kite system, whereby the relationship of the number of kites per ring depends on the predicted speed of the foil being used and the radius of rotation. Further parametric relationships describe such as the layering evolution properties of subsequent rings, ring area to kite area ratios, foil setting, tether routing, ring dynamic and panel cutting. Many other system parameter inter relationships have been explored. Designing kites specifically constructed to be faster and tethered from their dynamic root will also help.

    Where the lift line links to the rotary set, we used a rollerblade wheel to join the systems in the prototype to good effect.

    Investigation shows alternative configuration based on more rigid blades inflating their mount ring. Tip to tip tethering as shown is unlikely.

    Rotary to electrical

    Initial prototype power extraction was demonstrated by driving the kite set against a mountain bike wheel and braking against a squealing disk brake.

    For the initial “campsite and ecobuild” scale generation systems we intend to adapt an off the shelf e-bicycle regenerative braking system such as the Falco emotors Hxm2.0 .

    The regeneration characteristics of the synchronous motor can be easily programmed to adapt to the current driving characteristics of the kite set by using the existing falco software interface.

    An off the shelf inverter should be supplied to compliment the battery system.

    Larger scale grid tied systems have more stringent generation control needs and are not a part of the scope of this project. This integration of Daisy kite generator systems with grid systems will occur once a lower levelised cost of energy is demonstrated.

    Tethering, Anchoring and Weathercocking

    Multi point tethering provides safety through redundancy. The strength to weight ratio afforded by modern rope allows us to build lightweight tethering systems over large spans. It is desirable that failure of any component does not incur catastrophic system failure. This failsafe redundancy design principle includes the tethering of any single component. The Mothra arch and kite ring systems have incorporated this simple design safety principle.

    Driver kite tether drag can be reduced dramatically by appropriate bridling, ring layering distance reduction, inter layer tether setting point and appropriately doped tether selection. The goal is basically making less tether move through less air for the same energy transfer.

    As the wind moves so too should the whole kite system. This includes the anchoring system.The current prototype has shown itself to be capable of accepting 40deg wind shifts over the Daisy ring frontal surface without total loss of rotation or kite collapse. On initial trials the prototype demonstrated a single unattended inline lightweight parafoil sled lift kite can maintain Daisy to wind over a 3 hour test. However (without dropped base loop line backup) a single inline lift breaks the rule of redundancy. A kite stabilized to the sides of its wind window is a safer and steadier lifter, but the sides need able to weathercock. The weather cocking of spread lift kite tethering needs to be done at a radius greater than the ring kite tip radius. Several open solution proposals exist; ground belay sets set by operators, ring rope and carriage/trolley anchor sets, car anchor sets, rail anchoring, hex meshed lift kite upper ring weather cocking anchoring, hex meshed valley upper ring weather cocking anchoring, turret balanced kite bar booms wider than ring radius.

    For the initial camp / ecobuild scale model a simple spread anchored tethering of a Mothra like lift kite is proposed. This model has proven effective already in prototype testing. Tested stabilising results of our tethering system has been welcomed by open science group publiclaboratory.

    Software Architecture

    In order to maintain inherent system safety we decided to first prove the system without any onboard software necessary to run. However software systems have been invaluable in design and will be invaluable for the potential monitoring, control and communication benefits of the system.

    Grasshopper parametric algorithm 3d development language was used to model many possible system configurations before a likely looking prototype was chosen. Likely was based on years of kite experience and feedback from the yahoo AWE forum.

    To build the kite kite ring prototype, measurements were first taken from 1.3m ram air kites, which had approx ¼ span (half of the right hand side) removed. Once these sizes were set in grasshopper, the ring chute was developed to fit 3 driver kites. The ring was then divided into smaller parts in multiples of 3 until a neat pattern for bridling and kite mounting was found … and also so that each panel of the ring could be cut from a pattern able to fit on an A3 sheet of paper. Determination of the exact panel shape required used the squish function in Rhino… not an exact surface 2 axis bending tool… but good enough… it worked quite well I think.

    NASA machine vision systems are available for experimentation, as are track and control software from TU Delft, other open source kite tracking is also available. Their use has previously been tailored to optimization of kite trajectory in kite yoyo generation systems. It was felt this approach may be too complex for the simple needs of the Daisy system.

    Useful electronics and software can be applied to Kite sheeting power control feedback, Lift kite position monitoring broadcasting and regulation, ensuring top and bottom rpm matching to avoid hockled lines, extended and improved access to radio comms, Falco emotors generator charging control and performance feedback, acoustic analysis of line and bearing performance, inspection, safety lighting, lofted generation for localized operations, weather data, network broadcast availability, line tension data ... and more besides.

    For this project I plan to use easily accessible and replicable software platforms with reliable operating systems, simple programming, stock open source wireless solutions… yes old android phones will be repurposed to perform a lot of jobs.An old android phone with wifi tethering makes any kite a valuable new tool in it’s own right. Free software (such as RPMMeter) is available for rpm monitoring using either the accelerometer or the proximity sensor. Giri software can tell you the speed of your gear wheels with acoustic monitoring.

    Software demands of heavy duty controls such as would be used on a kite sheeting system with through axis relative tensioning will again need to be made failsafe. This feature will not be attempted in the camping ecobuild scale project. It is a feature very suited to a post and universal joint mounted model.

    3d system file models are available from or emailing rod directly

    Rod dot read at windswept-and-interesting co uk

    Communications Architecture

    Assuming that the lift kite has a good 3G+ data signal, It is logical to use this as the chief means of accessing kite data remotely. Software such as the “Wireless Tether For Root Users” app for android combined with phone state monitoring software such as stealthgenie can allow recycled phones to take on new uses.

    The Falco emotors generator has a proprietary dongle device for remote pc configuration. I think it’s Bluetooth . I’ll explore this further.

    File and Database Management

    Assuming this device proves popular, it would be best that people with sewing, design, kite, sales, installation and other interests would be able to share in the project.

    Details of possible partnerships for delivering a distributed open network enterprise management solution are currently being considered.


    Lift kites can be extremely powerful and are not to be toyed with. Thin lines under tension can cut like cheese wire or give really bad friction burns. Special precautions must be taken to ensure safe lift kite operations. Always have correct fitting protective clothing. Operators should always have an emergency rope cutting device available. Never leave a kite which is un doused or readied to launch. Stay upwind of kite surfaces. Keep control lines neatly laid and obvious.

    Once the lift kite is launched, prepare the Daisy rings upper end to be hoisted up to flying position. Hold and lift the handles fitted to the daisy generation wheel. Apply the Daisy brake. Hoist the Daisy back end to flying height. If all of the driving kites are set and open ready to drive, Release the brake. If some of the driving kites are collapsed to the inside, slowly release the brake until you can reach the collapsed driver kite, prime the driver kite (throw it outward) and when ready release the brake.

    Section still to be completed


    Hardware Detailed Design

    Software Detailed Design

    Internal Communications Detailed Design


    Interface Architecture

    Interface Detailed Design


View all 4 instructions

Enjoy this project?



Philip R wrote 11/12/2017 at 21:10 point

Hi Roddy - I was intrigued on (only just) seeing your videos. I'd been tinkering around on somewhat similar lines for a while - particularly with the basic idea of having a lifting kite separate from the rotational/generating component (ie. unlike Kitegen etc). I've not gotten as far as hooking anything up to a generator - just spinning a bike wheel. I was a bit stumped by the difficulty in finding/making inflated aerofoils which would greatly improve the power/weight ratio (and would, I imagine allow for scaling up - perhaps on a massive scale). Have you taken things any further since a year ago? I've got my own humble effort on Dropbox if you want to have a look. (I'm completely new to Hackaday so not sure about putting up videos here). Best wishes, Philip 

  Are you sure? yes | no

Roddy "Rags" Read wrote 09/25/2016 at 20:44 point

up to over 300w now with a 3 ring stack system...

  Are you sure? yes | no

Roddy "Rags" Read wrote 12/27/2015 at 12:19 point

I'm quite sure with the coming tweaks That I'll be getting a lot more out. Portability is the really cool thing, it's portable & it aids transport by charging up a bike battery... So it's even more portable.

  Are you sure? yes | no

andyhull wrote 12/26/2015 at 20:55 point

Excellent progress. 100W from a mobile wind generator. I'm impressed. What are you using as the generator? 

  Are you sure? yes | no

Roddy "Rags" Read wrote 07/29/2015 at 08:33 point

I've taken Daisy to the next level... in a literal sense. Here is a video of a 2 stack Daisy.

  Are you sure? yes | no

andyhull wrote 08/29/2014 at 14:31 point
What output power can you get from this?

  Are you sure? yes | no

Adam Fabio wrote 07/29/2014 at 05:51 point
Amazing project Roddy! Thanks for entering Daisy in The Hackaday Prize! This could be an awesome accessory for any emergency power setup, as well as supply supplemental power on a day to day basis. Don't forget about the 2 minute video requirement for The Hackaday Prize. Good luck - you might just be flying a kite all the way to Space!

  Are you sure? yes | no

Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates