[Ground-station] Sounding Rockets and FPGAs - at a University

[Michelle Thompson]

Where are the successful industry-university-opensource partnerships? It is
definitely not all bad news or misunderstandings. It's really difficult to
pull off collaborations given the very different goals and methods of the
(at least) three communities involved. Getting a better fit between
communities and getting better results does mean looking at what did or did
not work.

What does work?

Stanford, MIT, VT, and many other schools are way ahead of the game and
are, or have recently been, innovative in getting broad involvement behind
student work. Some of this appears to be 1) firmly enabling fundamental
research topics and 2) overfunding to the point where research risks can be
tolerated and 3) having a pipeline ready for productized
proprietary/secret work in order to keep the funding coming in.

None of these programs are perfect. But, relatively small investments have
resulted in enormous impact and results.

With the potential for many tens of billions of dollars of increased
spending on NSF coming up rapidly in the US Congress, it seems that working
well with "The Academy" remains and will be a very important option. Even
for smaller operations like ORI. Any increase in spending is going to go
through all of the many traditional mechanisms, such as NSF and NASA. These
are mechanisms that have produced an enormous quantity of work that, in
most cases, can be read and used by us for free. There's no way we can do
what we do without that body of work.

With the educational mission becoming more and more important to the
amateur radio service (and to the foundations that can fund ambitious work)
then it's going to be up to us to learn how to fit in and adapt, when
pursuing this sort of thing. That means understanding what the fundamental
research topics of a particular institution (or professor) are. Sometimes
this is easy to figure out, and sometimes it is not!

This is just part of the landscape and it's just part of what we do. We can
work on forward error correction and communications systems entirely apart
from academia, conferences, seminars, and student projects.

I believe that there's so much potential mutual benefit possible with "big
education", that there's an obligation to keep at it. The reason I'm
talking about it on the list and asking for feedback and sorting through
how this relates to wider open source work is to get our volunteers the
best possible opportunities.

-Michelle W5NYV




On Mon, Apr 5, 2021 at 11:05 AM Michelle Thompson <
mountain.michelle at gmail.com> wrote:

> This raises a point that keeps coming up over the years. And you inspire
> consideration about another issue.
>
> There's a huge amount of work/results in closed literature. Usually due to
> national security concerns.
>
> Part of the problem in doing open source work is running into people and
> volunteers that know something can be done, but they can't cite the source
> and do not feel like they can describe it. Missiles and rockets are
> certainly in this category.
>
> There has to be something that we can verify and test with the really
> remarkable work we are doing. We have some very good ideas identified so
> far. Let's keep trying to come up with ideas and clarify them as best we
> can.
>
> Based on what we've talked about, it seems that we have to "market" our
> ideas better to schools in order to get them to care. It's on us to make it
> "cool" enough for students to select. I'm completely biased and all-in on
> what we do, but full time people like me, and a 501(c)(3), and funding,
> have not been enough in the recent past to be "picked". I am actively
> trying to up the game here, any way I can, to get the work done and people
> educated. It's my fault for not knowing how to do this in a way that gets
> us recognized and engaged.
>
> -=-=-=-=-=-=-=-
>
> There's another subject here, and it needs to be brought up.
>
> Working things out from first principles, when it's necessary, can only be
> done in practice by people with the time, talent, and treasure necessary.
> In other words, by people with relative privilege. Or, by people that
> accept the burden and try anyway. I can think of several active projects in
> this category, some of which we support.
>
> And this sort of thing harshly limits the promise of open source work in
> two ways. 1) we sometimes really do need to liberate closed work and this
> is uncomfortable and requires some brass body parts and 2) it can most
> easily be done by people with economic privilege.
>
> This brings up how to pay for open source work. Not just for creation, but
> for sustaining engineering.
>
> I can try and describe this with words, but this image is better:
>
> https://xkcd.com/2347/
>
> We have to acknowledge this. I am not sure that we can address it, but
> it's a really big problem. If we can't solve it, then we have to accept
> some very unhappy limitations on open source work.
>
> We know and understand that all the major amatuer radio philanthropic
> organizations (in general) decline to pay overhead and salaries. They will
> pay for materials, parts, boards, and tools. But, they have rejected paying
> for overhead or salaries. Oddly, they will fund scholarships. The
> scholarship economy, the scholarship marketing movement, and scholarship
> funds are all about overhead and salary. And, privilege.
>
> First, scholarships fund the administration of schools. The reason we have
> a scholarship economy is because overhead and administrative costs are the
> things that have caused tuition to rise so sharply in the United States.
> This is a problem almost unique to the United States, and I apologize to
> readers outside of the US where this may not apply.
>
> Second, if you only fund college scholarships, and won't directly pay
> people to do open source work, then it's decided that the fund will support
> a system that filters out a huge number of talented people. As an extremely
> well-educated person, who went to an elite university, I am here to tell
> you that only funding college scholarships to "get technical work done"
> really does contribute to a homogenous tech scene. Technology is sexist,
> racist, and bigoted. If you are going "what?!" or want or need a good
> introduction to this problem, then
> https://www.npr.org/2019/03/17/704209639/caroline-criado-perez-on-data-bias-and-invisible-women
> is a great start.
>
> Philanthropic organizations that want to change the way things work on the
> ground, for real people, have to rethink their approach. If we only fund
> university scholarships and molecules (PCBs, components, etc) we are
> missing  almost all the leverage.
>
> I assert that if scholarships can be funded, then so can open source
> worker salaries. Open source worker salaries result in same or better civic
> good than scholarships. Direct payments to vetted workers might even be
> more efficient and will result in better economic outcomes to open source
> work. People directly affected by bad tech done by homogenous commercial
> interests? They are the very best people to pay to produce better work.
>
> Why is there resistance to paying for overhead and salaries?
>
> Because of a real and legitimate concern about waste and laziness and
> fraud and misunderstandings and completely different expectations and value
> systems.
>
> I'll give one example in the category of the work that is quite often paid
> for now, that was a complete and total mismatch.
>
> I was and am involved with an engineering project at a University,
> involving several large corporate partners. They had a certification
> program through a major online education app. This was a big open source
> effort. There was a paid (employed!) open source evangelist, traction with
> volume production for the hardware, and lots of truly great
> code/applications/creativity/Makers etc.
>
> When it came time to produce the instruction - that's when we had a big
> problem. These were classes that people were going to take online for a
> *certification* from a university. The deal was two professors, with full
> support of the university, took the entire amount of money to produce the
> educational content and... the money disappeared into the University. No
> content was produced. After an appropriate amount of time for content to
> appear, inquiries were made.
>
> Nothing could be done.
>
> Now, this was not what was supposed to happen. There was a contract, there
> were expectations, multiple meetings, lists, posts, commitments, community
> momentum, a contract from the online supplier, etc.
>
> The professors produced brief 1-minute or so introductions for the 40
> class sessions. They didn't provide expected instruction, they did not
> review anything, and they did not produce content themselves. Getting email
> answered was hard. They honestly thought they had provided full value.
>
> They took the entire amount for the entire program and expressed genuine
> surprise that there was any unhappiness at all. It was about 40 minutes
> total content for a very large amount of money. This was "they way it was
> done". Ranks were closed.
>
> But, the show really had to go one. So, volunteers organized by the
> project employee came up with the contracted educational content, got A/V
> support at additional cost, edited the classes, dealt with the educational
> content distributor's many requirements, handled the student forum, handled
> the customer service forum, did the labs, and updated the content as the
> platform and the framework used rapidly changed. For a while. With no
> compensation, because the entire amount for all of this work was consumed
> by the university up front.
>
> The only option to enforce some sort of sharing of the money was to sue.
>
> That option was not pursued. It was politically impossible for an
> industry-backed and industry-funded consortium to sue a university where
> three of the board members had graduated from. Total non-starter.
>
> The staff member and volunteer corps on this project were rapidly burned
> out. The people that were really good at this work, and wanted the project
> to continue, moved quickly away. No project like this is planned now or for
> the future. The university doesn't understand why the "community" isn't
> "more forthcoming" with "efforts". Well, it's because they all remember
> what happened.
>
> So, yes. There's reasons for declining to fund "overhead" and "salaries"
> when stories like this are not uncommon. I know many of you reading this
> have very similar stories of your own.
>
> However, a blanket prohibition hits that team in Nebraska (xkcd) pretty
> hard. They can't get the justice of compensation for their work. Yes, the
> project was given/published as a gift. But, as we know, a gift economy only
> works when there is reciprocity of at least respect, if not value. Lacking
> both, it collapses over time.
>
> If we want open source work to be enduring, then we have to start
> seriously considering paying people to do the work, instead of assuming
> only rich educated people will spend literally years of their lives putting
> up with the static to make great things happen. All endeavors succeed or
> fail based on how they can scale. We have a scaling problem in open source,
> and in open source funding, and none of us are immune from it. One of the
> ways we can solve it is "pay the people that can do the work".
>
> Larger foundations that have the money, and need to spend it, have a
> unique opportunity here to fundamentally change the world. Yes, work on
> university projects. But, please, seriously consider paying open source
> workers directly. There are so many people that would do this full time if
> they could. They've said so. They are not hard to find. A contract for 1-5
> years for specific deliverables is an easy legal document. The benefits are
> enormous. Right now, we have dozens of really amazing open source workers
> begging for dozens of dollars on Patreon and Kickstarter. Philanthropic
> organizations could fund all this work, today. That's literally just the
> tip of the iceberg. It's only the most privileged and most able open source
> workers that have set these sorts of things up.
>
> Back to sounding rockets - I'm going to move forward with some discussions
> as soon as some school deadlines are done. I am hoping we can take
> advantage of both Research and Development opportunities.
>
> More soon,
> -Michelle W5NYV
>
>
>
>
> On Sun, Apr 4, 2021 at 8:13 PM Thomas Savarino via Ground-Station
> <ground-station at lists.openresearch.institute> wrote:
>
>> I think that some meaningful work of integrating navigation and
>> communication is possible here.  I got the idea from some email where
>> someone said something about using the sounding rocket behavior for
>> something.
>>
>> I had two ideas for experiments
>> 1. Attach cheap accelerometers to crystal frequency sources and measure
>> the frequency drifts during rapid accelerations. This measurement could
>> provide an error signal somehow to correct the main clock frequency in a
>> system.
>> 2. It might be combined with something else, maybe a 3DOF Inertial
>> measurement unit that would measure accelerations and maybe autocorrect
>> Doppler shifts that occur during a flight.
>> I suspect that these ideas are pretty old in some areas, like ballistic
>> missiles, but getting it to work wouldn’t be that easy.
>> I think that some meaningful work of integrating navigation and
>> communication is possible here.
>>
>> So, there you go.
>>
>> Best
>> S
>>
>> Sent from my iPad
>>
>> On Apr 4, 2021, at 2:47 PM, Alex Wege via Ground-Station
>> <ground-station at lists.openresearch.institute> wrote:
>>
>> 
>> >>"Adapting to harsh and changing conditions quickly and reliably is a
>> big systems challenge for us. Is a sounding rocket the right entry point to
>> test this sort of work?"
>>
>> I think this is an awesome idea! As a recent graduate of the University
>> of Minnesota rocketry team I can tell you it's difficult to build a perfect
>> telemetry system for a rocket -- especially supersonic ones.
>> They might even appreciate just running some adaptive coding and FEC
>> blocks like in DVB-S2 to improve link stability (assuming they ran into
>> similar issues).
>> That would also be an opportunity to test out (by proxy) the dynamics of
>> our adaptive coding system in a stressful environment.
>>
>>
>> On a less related note, I think our system would be perfectly suited for
>> any high altitude ballooning teams to experiment with -- that would be
>> really cool to see.
>>
>> -KE0RYT
>>
>> On Sun, Apr 4, 2021 at 11:32 AM Jay Francis via Ground-Station
>> <ground-station at lists.openresearch.institute> wrote:
>>
>>> Michelle Thompson via Ground-Station wrote on 4/4/21 10:39 AM:
>>> > If we were to propose an FPGA experiment on a sounding rocket (this is
>>> > with a University), what would be the best experiment?
>>> It's probably easier to get FCC STA licenses for sounding rocket
>>> launches than orbital due to the limited duration.  I've done it a
>>> couple times now for S-Band telemetry on vehicles.  This could be a way
>>> to test very experimental modulation/protocols that might not be
>>> approved for orbital operation.
>>>
>>> Flight testing deployment of very small hardware (similar to Ambasat
>>> size or smaller) may also be possible since there's no orbital debris
>>> tracking issues.  It could be interesting to have a sounding rocket
>>> deployment mechanism to test swarms of small networked "satellites".
>>>
>>> Experiments flown on sounding rockets aren't necessarily only activated
>>> in space (unlike cubesats).  They can be designed to run through the
>>> whole flight.
>>>
>>> The flight environment (acceleration, shock and vibe) of most sounding
>>> rockets is a bit harsher than an orbital launch - be prepared for that,
>>> or take advantage of it :-)
>>>
>>> --Jay, KA1PQK
>>>
>>>
>>>
>>>
>>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openresearch.institute/pipermail/ground-station-openresearch.institute/attachments/20210405/97205e50/attachment-0001.html>