$50k fast funding program for low-cost science tools

Also eager to see @hikinghack’s hat on screen. :womans_hat: And thank you @davidtlang for the responses from you and your team on this matter, I’ve sent you an email to sign up for the call next week. BTW I’m a huge fan of OpenROV, thanks for developing an amazing success story for open source and open science hardware!

Pedantic question first: :timer_clock: You said 10:30 EST (UTC-5) in your message, but I thought the US east coast (presumably what you meant by the time) has been on EDT (UTC-4) since 12th March 2023??? If you can confirm, I’ll slot the meeting into my calendar accordingly.

Regarding the wider issue that @julianstirling has rightly pointed out, I was also considering submitting a proposal to work on the stereoscopic camera trap extensively discussed in this thread. However, upon seeing the applications that have been rejected, I am seriously discouraged from submitting something, as the camera trap stuff also don’t fit the hypothesis-based paradigm.

I share @julianstirling’s disagreement with the claim that framing - and possibly shoehorning! - projects as testable hypotheses will consistently result in “better” work. In my view, hypothesis-driven research is a very narrow subset of how science is done. One one hand, I happen to be working with researchers in medicine and clinical psychology at the moment, and can appreciate the importance of clearly defined hypothesis-based methodology in these fields (and tested with a double blind randomised controlled trail, etc.). On the other hand, I’ve worked across enough disciplines to know that this is not always the way to go. (Having a very multidisciplinary background can be a blessing and a curse, but in this case is a mind-expanding blessing!)

To add another example, I was part of a scientific team that studied the impacts from the 2010 Deepwater Horizon oil spill on marine wildlife in the Gulf of Mexico. Our work was highly exploratory with practically no hypotheses regarding what we’d find. “Despite” this, our results were published in journals such as PNAS, plus getting millions of USD for the research consortium. I have very deep misgivings and criticisms about how the scientific establishment deify and obsess over “high profile” journals and grant money. IMO that’s definitely not scientific! But I give these examples because even this ossified and byzantine establishment considered our work to be “good” enough.

The reason I used this example is that, in my humble opinion, the issue is deeper (and wider?) than whether development of hardware/instrumentation is “science” (and I believe it totally is in the way which @julianstirling, @naikymen, and @Harold described). It is about whether anything that doesn’t fit the rigid hypothesis testing mold counts as science (or “good” science). In my view, the answer is of course yes. There are various approaches for deductive and inductive enquiry that can fruitfully expand scientific understanding without an explicit hypothesis.

As @hikinghack pointed out, and also from my own experience, keeping things to a particular format is important from a practical perspective and totally understandable. More than that, I would even (possibly controversially) say it’s OK for Experiment to choose to only support science in the narrowly-defined hypothesis-testing form as described by @davidtlang, even if I disagree.

However, if that’s what Experiment chooses to do, then I suggest it should not in any way state or imply that the platform speaks for, represent, or support the huge diversity of ways in which great science could be done. Also, as @davidtlang et al. recognised (which I appreciate! :hearts:), the way the current low-cost science tool funding call is not well-designed and not a good fit for that narrow definition. Sadly, a huge opportunity cost has already been incurred in the form of the valuable time and effort that some of us have put into their submissions.

I think @hikinghack’s proposal is a good first step towards a compromise for the immediate problem at hand, and that a longer term deep reflection on exactly what Experiment is about and wishes to support would be useful.

The above represents my current views on the matter, which I’m happy to explain in the upcoming call if needed and/or desired. But I’d like to write them down here for the record. Thank you for your attention. :sweat_smile: :bowing_woman:

P.S. IMHO this thread is a good demonstration of the importance of having philosophy of science/theory of knowledge be part of scientific professional development/education. Some might not care (or not want to care) about such abstract concepts, but the problems demonstrated in this funding call show us the negative consequences of not thinking through (or not thinking enough about) high level/first order principles about science and what it is. I.e. great science such as the development of instrumentation not receiving the recognition and support it deserves!

2 Likes

Very well reasoned response Pen. I totally agree with you that in certain fields hypothesis testing is essential.

:100:

They can run their platform how they like of course. There are 100s of grant programs I don’t apply for because my work is outside the guidance of how they operate. The frustration voiced above comes from two key failings:

  • The grant explicitly asked for the type of work this community does. Proposals we put in accordingly. After taking the time to tailor a proposal to the challenge we were told that time was wasted.
  • Secondly, Exactly as you say Pen. If they came to the platform and said “Whoops, yeah our bad. We want tools, but our platform is set up for working a different way. We’ll make this clear ASAP”. But instead they came over and told is that if we adopt their narrow definition of science it will “result in much better projects”. A statement of sheer arrogance that really hammered home this is not about the administrative burden of different project types, but a value judgment on other ways of doing science.

Funding ambiguity

I think some of my response was strong because I believed that the money had been raised by leaders in our community to fund our community. That “Experiment” was just the platform that they chose to distribute this money. As such Experiment standing between our community and money it has raised was abhorrent.

However, I do realise on closer inspection that while @jcm80 and @shannond are listed as science leads. It was just my (probably ungrounded) assumption that it was the Open Science Hardware Foundation that got the money from Schmidt Futures, and put it through Experiment.

I now realise that the Experiment Foundation that @davidtlang mentioned is organisationally different from the platform, despite sharing almost the exact same name. See About — Experiment Foundation. It seems that Experiment Foundation are the foundation that facilitate the Challenge Grants, though there is no mention of that in the challenge grants page! Organisational ambiguity increases as David’s email domain “experiment.com” not “experiment.foundation”.

If it is Experiment (in either organisational form) that got the money from Schmidt, this is a different issue. Experiment then have the right to attach whatever value judgements they want to applications as long as that was made clear to Schmidt when they got the money. If Schmidt Futures think they are funding low cost science tools not exploratory research, then again our community are being stolen from, the money was assigned to us and is being diverted. I do not know American Charity law, but as the Chair of Trustees for a UK charity I know that I cannot tell a funder one thing and do another in practice.

I think what has to be answered in the call is:

  • The route of these funds and were their raised specifically for low cost science hardware, or were they raised for exploratory research?
  • What are the separation between:
    • Platform
    • Foundation
    • Science lead
    • “supported by” - sponsor
4 Likes

Hi Julian

I’m happy to address any ambiguity about the relationship of the Low-Cost Tools for Science challenge grant to the GOSH Community. Simply, there is no association with the GOSH Community or Open Science Hardware Foundation, beyond an overlap in individuals who are involved. This is why neither are mentioned on the challenge site or in the original email from Shannon that started this thread. If we had raised $50k ourselves we would have been quite pleased and definitely mentioned that :wink:

The challenge is an initiative of the nonprofit organisation Experiment Foundation, who independently conceived of the idea and fundraised to make it happen, resulting in a grant being made to them by Schmidt Futures (for which kudos to @davidtlang and colleagues). It runs on the experiment.com platform, which means projects have to follow their guidelines. It’s one of many similar challenges for ocean research, conservation, cellular agriculture and more that have a range of different sponsors/funders.

Shannon and I were later invited to act as science leads for the challenge in our personal capacities, not representing this community or any organisations. The role of science leads is to review incoming applications and make funding allocations after they have successfully passed through experiment.com review, which is the process that is being discussed in this thread.

I hope that clarifies things!

Jenny

5 Likes

Correct! EDT UT-4

I look forward to answering questions about the program. Please send me an email – david@experiment.com – if you would like an invite to the meeting.

To reiterate: Both I and the Experiment team consider tool-building to be a first-class scientific effort.

Here are examples of tool-centric projects that the Experiment Foundation has funded through this program:

I will do a short background on Experiment, the Experiment Foundation, and the funding programs that we’ve been running for the past year.

-David

2 Likes

Thank you @jcm80 for the clarification, it makes more sense now!

And thanks @davidtlang for confirming the time. I’d like to clarify that, IMHO, the issue is not if Experiment supports the tools with which we do science, but rather how that’s done and the limitations of that model. In any case, looking forward to the call this week and for all to learn from each other!

Anyone else want to join the call tomorrow? @julianstirling can you make it? If not, I’ll make sure to address the questions you listed.

1 Like

I should be able to make it. I sent an email to you on 5/5/23 (23:47 BST). I’ll pop you a DM on here.

For those who like me hate time zone converting, here is an event time announcement is set to @davidtlang’s location* that should have a time for your area.

*Not sure where on the East Coast David is, so I picked DC. I feel it is appropriate to just pick the capital after living in the US for years…most Americans had no idea that not everyone from England is from London.

1 Like

I’d like to join. Sent you an email to get the link
–a

1 Like

Hi all,

Nicole at experiment has replied to my email above. I have been asked in the interest of openness to share her response, just as I shared the previous messages:

The links are:


Or course, I never really shut up, so here is my response to Nicole:

1 Like

Hi everyone!

I won’t be attending to today’s meeting, but I don’t want this to be interpreted as lack of interest. I also wrote a message to Nicole, and no response has yet arrived.

In my opinion, the idea that hypothesis-driven projects are “more successful” is mostly a spurious correlation. It may be simply be that projects more likely to succeed have well-defined objectives and timelines.

The instrument we’d design and build would be tested and used also for my PhD, for answering questions about decision making processes in a molecular biology model. My research question can be stated explicitly, but I have not been given feedback on how to improve the draft (nor was I aware that it was in draft state and could be “republished”), nor did I realise it was relevant to the application.

I hope you have a productive meeting. :slight_smile:

Best!

1 Like

Meeting happening now here:

We had a productive meeting. I was an hour late because I am an idiot, and messed up time zones despite the link I sent. I think a recording will be shared soon.

I think for me hearing the context from @davidtlang of the grant really helped. Like always in written asynchronous discourse it is easy to misunderstand each other.

I think we have a way for instrument driven open hardware projects to be funded, we just need to show that we are actually discussing the context of the scientists that they will help, and how we will demonstrate the tool is fit for that purpose. This seems very reasonable to me.

I need to watch the rest of the video when it is online to see the bits that I missed due to time-zone confusion.

1 Like

uff, I read through the whole thing asking myself if I should drop the new iteration of the #ASKotec Kit [1] in modules, starting with a toolbox based approach together with a more scientific kit. In the essence making the big case go into smaller variations for different use-cases but still based on the field experience we have gathered within the https://asknetc.community context.

After reading the conversation I’m really hesitating to take all my free time trying this, as even more clearly science focused projects where rejected at first. Probably finding a proper hypothesis is somewhat possible but I’m really worried that might be the wrong direction and I can save that time for another education oriented application.

So in advance thank you for that open discussion here, that helped clarify the original goal and direction of this grant, though I agree with @julianstirling this might not necessarily be in synch with reality when we talk about “low-cost science tools”. So maybe apart from some more guidance (Examples helped) what the “Science” should be focussing on, at some point you might consider the part “Building the tool- and skill-sets to build the tools” as a valid option too :wink:

[1] GitHub - opencultureagency/ASKotec: A mobile tool and resource kit as a makerspace in a box

1 Like

Hi all, and thanks for sharing Julian.

I’m not sure I got the meaning though. Does it mean that we should rewrite the application to tie the tool to a specific experiment which would have to take place happen in the context of the award?

Would the projects mentioned by David/Nicole be good examples of this?

Best
N.

It seems that they are happy for it to be a two way conversation with Experiment as we prepare. I think we should be able to speak to them as we iterate to something both Experiment and the applicant are happy with.

I am hoping I can still lead with the hardware project, and the fact we want it to be open and anyone will be able to build it. As part of this we can give an example of one of the key scientific applications of the instrument, and do an experiment to check the instrument performs this task well.

1 Like

Fantastic! Has a preferred communication channel been suggested? I don’t seem to be able to republish the draft on the experiment.com website.

PS: Sorry to be asking so many questions after missing the meeting. I’d be glad to watch the recording when it is available.

Perhaps make a new one? Or email Nicole or David? You could also ping @davidtlang on here.

1 Like

Hi @hikinghack @hugobiwan
According to my opinion the project is really cool.
For scientific application, i can imagine all the application that need touch sensitiv documents, i’m thinking of :

  • automated generation of map from openstreetmap
  • touch sensitiv transcription of painting
    And all the application in pedagogy for visualy impaired students.

Best regards

2 Likes

Hi David @davidtlang I was wondering if there are new guidelines or advice for updating the applications.

I wrote to Nicole several times but have not received a response.

Best,
Nico

1 Like

I know Denny and Nicole are busy working through reviews and emails today! A lot came in over the weekend.

4 Likes