blog post: The Future of Wind is Airborne

I recently wrote a blog post breaking down the airborne wind energy industry for the general public and since I am not a technical expert, I would love to hear if I missed anything or if I got anything wrong to get a better understanding.


I think the article is very nice and spot on

Maybe mention cubic scaling of mass vs square scaling of wing area…



Sorry I am not aware of this, why is it the case that there is cubic scaling of mass vs square scaling of the wing area? Is that mostly because you need to use stronger/thicker material when you increase the size of the wing?

Rigid wings undergo more or less cubic scaling of volume and mass.

Flexible wings also undergo scaling of mass, but at a lesser magnitude.

Page 532 of the pdf below:

The development originated as an Advanced Concept Technology Demonstrator research program from Natick Soldier Systems, whereby iteratively heavier weight requirements were levied (0.25 ton, 1 ton, 2.25 tons, 4.5 tons, 13.5 tons, and finally 19 tons). The wing sizes were 36 m², 102 m², 250 m², 350 m², 900 m², and 1,040 m², respectively. After a given weight could be reliably flown and landed precisely, a heavier requirement was set. The technology evolved to be modular, stackable, and scalable, using the root airfoil throughout. The 1,040 m² wing, made possible by modularity, was simply a swap-out of wider wingtips from the 900 m² wing. It is noted that as the wing became larger, a heavier wing loading was used.


Hi @daviddemeij
Yeah I’d seen your article flagged up for me on twitter.
As I replied there
From the article

As the kite gets larger and heavier it can be harder to manoeuvre and generates relatively less power (part of the energy goes towards counteracting gravity).

That’s 1 reason we use a network array scaling approach

Well done on your research

1 Like

Btw if you want to work on kites and AI, I have a few ideas for you :wink:


And let me know if you want any info on kite turbines.
They don’t need a controller and have continuous rotary output from flying loops right inside the power zone.

Hi @Rodread, thanks for your comments. I had a look at your website before, does seem like a very interesting concept with a lot of advantages! What would you say are the biggest disadvantages or challenges of this approach? When do you think you will have a 10 kW system deployed?

Also really cool that you open-sourced kite turbines <1kW it seems like this has a lot of potential for powering remote villages in the developing world.

I am curious about your ideas!

Thanks for the info, I included this in the article :+1:

Thanks @daviddemeij
Good Question
Biggest Disadvantages and Challenges for Kite Turbines

  • Proving reliability - The autogyro rotor has worked well in gusts according to the tests I have flown but the transmission is vulnerable to torsional compression and can break if you treat it too roughly. This will be mitigated with a generation control based on turbine positional and rotation data

  • Rarity - Hardly anyone is trying it. There’s hardly any acknowledgement it exists even in academic texts. This is the hardest thing for me. I think operating from a lower altitude doesn’t give theoretical operators such a hardon. Either that or the novelty and dynamic is too hard for them to model competently. It’s infuriating because for me this is the best solution for scaling, efficiency, safety, cost…

  • Launch and landing handling - We normally stretch out the turbine network and transmission before launch. This needs space to operate. So do all AWES. Extending the transmission after launch for higher altitude operation would be possible but I haven’t tried it. Same with adding rotor layers. We’re working on some other launch patterns and devices.

1 Like

@Rodread Thanks for the extensive response! I noticed that this approach is indeed quite unique and not pursued much. I added a bit in my article on this, let me know if I got anything wrong.

I think I got a decent understanding of the advantages and challenges of this approach and it does seem to have a lot of potential. I do see how it is hard to model such a complex stochastic system with so many moving parts. I love the networked array design, it looks like something that could have been designed by nature through evolution.

I am curious if you are willing to share. How far are you in developing a 10 kW system? It seems that your company is still very small, are you planning to grow the company or are you maybe planning to license your technology to a bigger company? Are you thinking of raising an investment?

1 Like

Thanks for adding us to the report @daviddemeij :smiley: :+1:

We’ve started phase 1 of a project to define, design and report on an automated 10kW system. Phase 1 is scoping - How we will assess the business, the turbine, How we will test, what we are looking for in automation, What we’ll build, collaborate, regulate, what we’ll specify, how we validate, and How we’ll run the tests…
Just me in the office this week. NEED to grow the company with investment. Failed again to get EEF & InnovateUK funding recently. (Too many European consultants on the plan this time - #Brexit yay) I am bailing the company out again with my kids toy money. Hopefully with the state of the NHS I can sell a kidney soon

1 Like

I know how that feels, I was helping a renewable energy startup on an EEF application for last year aswell and they were just a few points short to get the grant.

You could consider doing a crowdfunding campaign on a website such as Seedrs.

1 Like

I agree with Rod about AWE reporting bias and this is particularly true with respect to a cable drive system like Kitewinder which has been commercialized on a small scale. This is a true steady state system capable of operating at very high altitudes. I believe this system is ripe for scale-up.


There is a lot of figuring out where you are and flying the kite, maybe optimizing production. Interesting for me right now:

  • an automatic pilot able to keep a kite flying safely then land, in any conditions. This would act as a generic fail-safe able to recover from any exceptional situation (assume kite is working perfectly)
  • estimate the actual wind gradient from flight

I was hoping to present some of my progress on the latter at AWEC 2021 [held in 2022]

1 Like

@tallakt Regarding this first point, what do you think about Skypull’s approach? They claim they can still safely land the kite even if the line breaks.

To train an AI to always fly the kite safely and land in any condition will probably require a realistic simulator that can simulate edge cases.

That simulator is not that hard to make. Conditions for a kite are very deterministic. Only wind conditions change. No other actors, video feeds, difficult planning etc required

1 Like