Re: Request for comment before BONSAI 2020 General Assembly #communication #dataliberation

Brandon Kuczenski
 

Hi all,
I agree that the USEPA work under Wes has been really impressive and broad. Compared to BONSAI though, it is both very well capitalized and very constrained institutionally. They are (I think) limited to incremental changes that are not especially transformative. I think they are right alongside BONSAI in terms of trying to figure out how to meld their work with others into a new paradigm. I think this is going to be a process of slow accretion, where the key contributions will provide cross-links between different projects.

Based largely on my experience in the 2019 hackathon, I see BONSAI as a sort of "incubator" or synthesis lab. Taking problems that can be observed and solving them in extensible ways. The ReCiPe 2016 implementation- that is really impressive and a great example (and this is the first I'd heard of it). LCIA has to be redundantly implemented by all the softwares- and even Ecoinvent has to do it- because there doesn't exist a consistent, stable third party version. Now there does.

One thing you could do that would cross-link would be to provide a static export in the EPA's (quite simple) LCIAFormat (https://github.com/USEPA/LCIAformatter/blob/master/format%20specs/LCIAmethod.md). Not sure if it does this already, or if there is a generic export from brightway into a static format. Chris, Didn't you also come up with a serialization format in your regional LCIA roadmap, using data packages? I am all about serializations, especially of shared open resources in simple, transparent formats.

I also agree with Matteo's comment about "integration" as opposed to aggregation being a valuable thing to pursue. Based on the slide that you posted, I don't really see how CDLCI is contributing to that- it seems like they are just modernizing their own integrated pipeline and trying to get their partners' offerings to pipe through it, rather than to cooperate or equally interoperate with it. But I'm just biased against PRe as "old guard."

As for engagement- that is not limited to LCA- I have noticed that it afflicts all of industrial ecology. Look at the R community and especially the "tidyverse" (e.g. for a community that has really embraced novel communication forms to build a truly online collaborative network. UCSB is near the epicenter of that movement (e.g. https://eco-data-science.github.io/ ) and I still feel sharply to the outside of it. But then I don't use R. People need tutorials and hand-holding and blog posts and such that will help them:
 * make websites
 * publish their code
 * document + test their code
 * write tweets / respond to tweets
etc. I personally need help with this. I found cmutels' python-skeleton to be really helpful in getting my very first package published on PyPI (https://pypi.org/project/synonym-dict/). BONSAI projects (and brightway-derived projects like the activity browser) are the only place where I've seen true open-source collaborative development in LCA and that is potentially a big deal. But it does take a lot of effort and a culture of constant engagement to maintain it. 

Some people, e.g. the "energy twitter" community, are good at this. I am not. I have found that I have a tremendously high activation energy and fear of rejection / inadequacy / impostor syndrome that prevents me from putting stuff out at all. Rather than a "hackathon" I could imagine a periodic meet-up event, e.g. weekly on slack or similar, where we all get our jekyll sites up, write tweets / blogs and then re-tweet each other, until it becomes more of a habit. Writing documentation could also be done this way. Writing clearly-stated problems, competency questions, demonstration projects, are all valuable additions.

Just some thoughts. Keep up the good work.

-Brandon



On Tue, Jun 23, 2020 at 8:56 AM Bo Weidema <bo.weidema@...> wrote:
I echo Matteo's comment :-)

Bo

Den 2020/06/23 kl. 17.40 skrev Matteo Lissandrini (AAU):
> My comment is the following:
>
> The real value proposition that none of the existing resources seems to have is the possibility to *integrate* multiple data sources.
> *Integrate* is complementary but very different from *aggregate*.
>
> Resources like :
> https://www.globallcadataaccess.org/
> are not dissimilar to
> https://datasetsearch.research.google.com/
>
> are aggregators of data. You can find a dataset, but then you are on your own in determining if the dataset contains the data you need, has the correct format, is coherent with your model, adapt it to your workflow, etc.
>
> What I see on GLAD (but I only read the homepage) seems more towards the *integration*, which is good, we should keep a close eye on that.
>
> I think the BONSAI with the Ontology and Mapping resources is a step ahead in the *integration* direction.
>
> I think the next step is documentation,  from LCA domain experts interested in integrating data (BONSAI)  for LCA domain expert interested in sharing data (all the providers).
> As we have a lot of tooling and work done and we just need to show people how to contribute.
> With more contribution we obtain more use-cases, we bullet-proof the workflow, and we improve.
>
> Yes, documentation is boring, and most of us don't feel are achieving anything if we are not writing code, calculating numbers, or running experiments.
> Yet, I see this as the crucial part for "lift-off", otherwise our space shuttle will always stay in the hangar redesigning the engine for N-th time but never turning it on.
>
>
> Best,
> Matteo
>
>
>
>
>
>
> ---
> Matteo Lissandrini
>
> Department of Computer Science
> Aalborg University
>
> http://people.cs.aau.dk/~matteo
>
>
>
>
>
>
> ________________________________________
> From: main@bonsai.groups.io <main@bonsai.groups.io> on behalf of Benjamin W. Portner via groups.io <benjamin.portner=bauhaus-luftfahrt.net@groups.io>
> Sent: 23 June 2020 16:58:26
> To: main@bonsai.groups.io
> Subject: Re: [bonsai] Request for comment before BONSAI 2020 General Assembly #dataliberation #communication
>
> My thoughts:
>
> - BONSAI and other projects (Electricity LCI<https://github.com/USEPA/ElectricityLCI>, US EEIO<https://github.com/USEPA/USEEIO> LCIA base data<https://github.com/USEPA/Federal-LCA-Commons-Elementary-Flow-List>, LCIA data formats<https://github.com/USEPA/LCIAformatter>, CDLCI project<https://www.pre-sustainability.com/news/harness-the-power-of-the-lca-community>, GLAD<https://www.unenvironment.org/explore-topics/resource-efficiency/what-we-do/life-cycle-initiative/global-lca-data-access-network>, IEDC<http://www.database.industrialecology.uni-freiburg.de>,...): I agree that it is hard to keep up with the numerous projects out there. We should definitely avoid reinventing the wheel. The question is: How? Adopting existing data, formats, etc. is good. But how can we do that if we don't know about them? Do we need some sort of "radar" for new projects and a "screening system" to filter out the portions of those projects, which are interesting for us?
>
> - Community engagement: I agree. I don't know if the existing BONSAI technology can help. Would be awesome if it could! If not, we need to find better ways to engage the community. Pré's LCAList is a good place to find advice but it is a poor place for collaboration. GitHub is a great place for collaboration but I don't think many people from the LCAList are there. I wonder if there is a way to bridge the gap between both communities.
>
> - BONSAI's 2nd work track: I was originally in favor of this idea but I lost momentum because I felt like support within BONSAI was weak. Furthermore, I feel like the release of GLAD made this the work track obsolete (although I might be wrong). I think there should be a new discussion about this.
>
> - Hackathon: I really regret that there is no in-person hackathon this year. For a newbie like me this would have presented a great opportunity to get to know everyone. Furthermore, I was planning to use it as an opportunity to get acquainted with the work previously done in BONSAI. I agree that an online hackathon is a poor substitute. But I think that an online hackathon is better than no hackathon :)
>
> That's it from me. Looking forward to hearing the results from the GA.
> Ben
>
>
>
>






--
Brandon Kuczenski, Ph.D.
Associate Researcher

University of California at Santa Barbara
Institute for Social, Behavioral, and Economic Research
Santa Barbara, CA 93106-5131

email: bkuczenski@...

Join main@bonsai.groups.io to automatically receive all group messages.