In the spirit of transparency, we will note that the author and editor do some volunteer work for the OpenSustain.tech team. Our bias is very clear, we think.

OpenSustain.tech And Other Ventures

Tobias Augspurger is a passionate advocate for harnessing the power of open-source technology to tackle environmental challenges. As the prime mover behind OpenSustain.tech (though he insists he does NOT want this to be a one-man show and kindly requests help from anyone who may be interested), Tobias has worked tirelessly to connect, catalog, and amplify open-source projects that advance sustainability. 

With an engineering background rooted in instrumentation robotics and a commitment to fostering transparency, Tobias envisions a world where open collaboration drives environmental solutions, curbing greenwashing and fostering innovation. His remarkably forward-thinking vision resonated with this pair of tree-nerds turned-data nerds. 

A regular speaker at events like FOSDEM and LF Energy, Tobias combines technical rigor with a deep commitment to democratizing access to impactful environmental tools. His methods of discovering and promoting projects—from scouring GitHub for hidden gems to distributing stickers at conferences—showcase his creativity and tenacity.

Tobias’s vision for OpenSustain.tech is bold: to create an environmental operating system composed ecosystems of existing tools and data that are not only open but also truly accessible and useful to both experts and "civilians."

His dedication shines through initiatives like the creation of automated news feeds for open sustainability and exploring how AI could supercharge environmental problem-solving. Tobias’s work has already laid the groundwork for more transparent, impactful, and collaborative approaches to combating climate challenges worldwide.

In addition to OpenSustain.tech and ClimateTriage.com, Tobias and his colleague Michael Scheiwiller have recently announced another exciting project: OpenClimate.fund, an exciting initiative developed with ecosyste.ms/ to financially support open-source climate and sustainability projects. Open-source fuels thousands of solutions, from monitoring forest health to crafting eco-friendly designs, and this fund aims to ensure developers have the backing they need to thrive. 

Both of these projects serve as "ecosystem services", supporting thousands of open-source projects in all environmental fields and providing the basic support for open source projects to flourish. New participants join the projects and get basic funding to cover running costs such as travel to conferences or hosting.

OpenClimate.fund simplifies the process: contributors discover and nominate impactful projects, and the fund provides financial support to bolster their work.

The best part? Anyone can choose how the money is spent! If you vote for a project, it will be awarded €100. With an initial funding round offering €10,000 to 100 projects, this initiative sets a new standard for supporting open innovation in sustainability.

In our conversation, I spoke to Tobias about his platform, his vision for the role of open-source in tackling pressing environmental challenges, and how initiatives like OpenSustain.tech, ClimateTriage.com and OpenClimate.fund could shape the future of sustainability. 

We delved into the intersection of technology and forestry, the importance of transparency to combat greenwashing, and the creative ways open-source projects are already transforming how we address climate issues.

Tobias shared his insights on the challenges of engaging non-experts, the innovations emerging from open collaboration, and his hopes for a more interconnected and supported ecosystem of sustainable solutions. 

Interview

How does OpenSustain.tech ensure that the data is not only openly accessible but also understandable to non-experts? 

In the past, we have tried to communicate the value of open source software and data to the masses through a variety of campaigns:

The first report we produced is perhaps the best attempt so far. Given the very limited resources and time we had, it is by far not as good as I would have liked it to be. We also tried to make it accessible and in-depth at the same time, which made it very difficult to read at some points. The presentations of the report at various conferences were able to reach a lot of people in the open-source community.  

The Open Source Software for Climate podcast hosted by Richard Littauer was good in itself, but I had to pay 3000€ out of my own pocket for the production. The search for sponsors for further episodes of the podcast has so far failed.

I still think it's the right way to give the people behind the curtain a voice creating the key digital infrastructure to understand, measure and protect our natural world.

We are planning to make the podcast part of the insights we want to create for our next report. There are too many podcasts being produced at the moment and not many people take the time to listen 30 minutes.

Writing smaller blog posts about our unique insights and distributing them via various channels reached the most "civilian" persons so far. In particular, the folks at the Open Source Initiative have given us the space to talk about our work through OpenSource.net, which I'm very honoured to do. 

I have done several science slams in the past to communicate our findings to a broad German audience. Although I have always received a great deal of support for this, it has unfortunately not led to anything more.

We even tried to find an artist who would use all the images we had collected from over 2000 projects. Although many of the images show very brinsant and from my point of view artistically valuable content, the interest of "The Art of Open Source Software in Environmental Sustainability" remained low.

At FOSDEM 2024 I gave out 200 ClimateTriage.com stickers, which was a very interesting experience because again the feedback was great and I had very intense discussions with a lot of very cool people.

Have you found difficulty bridging the divide with the “civilian” world?

The problem is that non-experts generally have a misconception about open source and software development in general, not realising that open source is at the core of almost all software products we use today.  Understanding why numbers like a CO2 price require a huge amount of software, and why "open modelling" is so essential to it, is unfortunately beyond the attention span of many people I have spoken to.

Most people are not aware of the fact that open-source sits at the core of almost all software products we use today. Explaining that numbers like a CO2 price require a massive amount of software and "models" in general is also not so easy. Most people also do not understand what scientific modelling is and why you need it for a CO2 price.

What are the future plans to expand the reach and impact of OpenSustain.tech?

As we have not received the hoped-for donation of $10,000 from GitHub any money/donations/grants at the moment we do not know exactly in which direction we are going. At the moment we are considering building the following in our community:

A second report focussing more on the use and measurable impact of open source projects. This time, we want to build on the extensive metadata from ecosyste.ms to analyse the driving forces behind open source projects. 

A transparent and automated fund that distributes money across the environmental open source ecosystem. The decision on how to distribute the fund should be 100% based on the metadata we collect using open science methods.

An automated news feed of all new open source software and datasets in environmental sustainability.

An LLM that helps you combine the ideal open source projects, APIs and datasets for your environmental issue. Such an AI could help people combine the right open source tools for their environmental questions.

Since such a development would bring many unknowns, are very expensive and LLMs are also considered to harm the open source movement in general, we will most likely not go in this direction. LLM are also not really open science as their internal functions are not traceable.

How has OpenSustain.tech expanded since you first conceived of it?

The dataset is much better and more comprehensive, covering almost all topics, organisations and projects that are active and documented. Thanks to the help of Andrew Nesbitt from ecosyste.ms we could find a lot of projects via AI and other methods. It takes me sometimes hours of research to find high quality project that are missing, which indicates a high coverage.

It takes me sometimes hours of research to find high quality projects that are missing what indicates a high coverage.  This collaboration also takes the metadata we have collected over the years to the next level.

Are you surprised by how successful it has become?

Yes, absolutely. I'm still surprised that so few people and organisations are analysing and promoting "open sustainability" and the use of open-source / open science for environmental solutions in general.

I really wonder why open-source isn't as prominent in the climate and sustainability business world as it is in other areas like robotics, where I used to work in open-source. I still do not have an answer why this is the case or why open-source is not discussed much more in movements like Fridays for Future. 

One problem with open-source in general is that you don't know how, why, and who is using your project. Some people have cited our project, but as with most open-source projects, people do the same thing: Use it and move on. As with many services that are provided for free, only we have no advertising behind it, so we have no income .

How did you find all of the projects? Was it simply a matter of endlessly searching GitHub? 

So the methods I use to find projects have changed a lot over the years. I already had some good experiences when I created the Awesome-Robotic-Tooling list as part of my research and development work at DHL. Even without maintaining this project for 5 years now, it still gets more stars per day than OpenSustain.tech, which shows how little attention open source in climate and sustainabiliby gets. 

I also star every project I review so that I can easily see if I have already reviewed it. Stars are my GitHub bookmarks.

The challenge with anyone else trying to help curating the list is that they do not know what is already on the list and have to double check each project. This is very time consuming and the reason why it is not easy to help me with this.

As part of my work/study in this field, I have read journals and attended literature seminars where people present good papers on climate change, energy, environmental science and earth science. When they mentioned a keyword or topic that I did not know, I did a search about it. 

I looked for glossaries like these:

NASA's Global Change Master Directory (GCMD) Keywords

Wikipedia

Earth Labs System Science Glossary

Enel Green Power Glossary

European Environment Agency Glossary

IPCC Climate Glossary Research Demo

Follow people on LinkedIn, Twitter, etc. who are relevant to the field and check out their GitHub stars. This is how I found a lot of keywords I did not know about.

We also listed Taxonomy and Ontology for various field but I never used them to find keywords.

I check from time to time some Forums that are relevant in this space like ClimateAction.tech or openmod.org.

Andrew tried some embedding / AI tooling to find projects based on the topics / oneliner / README. Similar to this here. He can provide more details how he has done this. In this way we found more than 100 projects. 

In your 2021 talk with LF Energy, you say, “openness is the best way to reduce greenwashing.” I wonder if you can give more specific thoughts on this? 

Because openness is a way of demonstrating sustainable intentions through the use of open science and open source, demonstrating why something has been judged to be sustainable. By showing the math in form of code+data why you consider something sustainable you prove sustainable intentions. This allows that your assumpution, scientific models and data can improved and validated in the long term. All greenwashing is based on obscure and opaque figures that cannot be validated. If companies were forced to prove their claims using open science methods, we would significantly increase the barriers to greenwashing. At the moment, almost any green claim can be made about a product without any evidence. Hopefully this will soon change, at least in the EU. It is important to note that open science, the usage of open source + open data for scientific work, will not provide you good numbers and "right" statements about environmental impact from the beginning but it will provide the possibility to improve and validate assumptions and numbers on the long term.

How have you seen the “Unix Philosophy” being applied to sustainable technology since your platform launched? 

It is difficult modularity, a key aspect of open source projects as described in the “Unix Philosophy” and how much of the projects are recombined.Our report showed multiple projects that are reused a lot and we have the download numbers showing which projects are being reused as part of another software project.

Our next report will hopefully provide more insights how and where within the ecosystem modular open source projects are recombined and reused, but this demands analysing data science projects that are released with scientific publication. We are currently investigating data mining methods to automatically scan scientific papers for the software being used. 

Have any new innovations been created as a result of your list? Have any new innovations been created as a result of your list?

I really don't know because the feedback we get is very limited. I know of several grants that use our dataset to check what has already been done and where there are gaps. A lot of people have used our dataset to check what projects are missing. The Linux Foundation's sustainability strategy was also defined last year using our dataset.

Measuring the impact of OpenSustain.tech is very difficult because we do not track our users. I know of many areas, such as the cryosphere, where we suspect that we have helped several open source projects to find each other and build larger applications together, but I cannot prove it. In the end, for many people we are just a one-off service to help them find the right projects and organisations for the area they are active in. I would love to know how much more collaboration and new development we create, but I may never know.

How has marketing for open-source products evolved since you began this project?

Nothing has changed; in fact it has gotten worse. A large part of our reach was created via Hacker News and Reddit, but unfortunately the interest in our content there has waned considerably. In the past I have posted other climate relevant open source projects there with really amazing results like the Heat Pump Monitor. Unfortunately it seems to me that the wider hacker community has actually become less receptive to such content.

One of the major setbacks for the marketing of open source in general was that OpenSource.com was virtually shut down by Red Hat. The whole community had to rediscover itself on OpenSource.net but it will take years to create the same reach and content.

In your 2021 talk, you alluded to a continuous reforestation project. I am also interested to learn more about this (We are former reforestation contractors).

Here's the repo.

The problem with this project is easy to explain. No reforestation project was able to provide feedback via an API when and where a tree was planted. That's why we had no way to validate if a tree was planted and how well it was growing. That's why we could not use Open Science methods leading to the stop of the project. Selling numbers without the ability to track them and provide uncertainty is exactly what caused the carbon offset frauds uncovered last year.

In that video, you also called reforestation the only carbon capture technology that works worldwide and is efficient enough. I wonder if I can hear more about your thoughts on this? 

That's because trees with intact biodiversity around them are much cheaper than other direct air carbon capture methods that rely on man-made machines. Tree planting also stabilises the carbon cycle on the long term and direct air carbon capture machines need a massive amount of green energy and a lot of maintenance.

One of the innovators you have alluded to in this sphere is Digital Humani, who are also from Canada. Are there any other players who have emerged in this sphere?

We have seen some great projects coming up last year that should improve the numbers and data in reforestation:

Allometric

UDef-ARP

The Global Canopy Height Maps 

How do you ideally envision the future coexistence of the open-source tech and reforestation movements?

I hope that, as in climate science or energy system modelling, a toolset of established and accepted open source projects will be developed. Such a toolset would create standards for how good reforestation figures/data and plots are produced and validated. Such a toolset should be developed in an open community, combining the best tools, data and APIs. A start would be to create an open source reforestation forum or conference to facilitate collaboration. 

Another way to increase the collaboration between different project is to create video tutorials and courses together solving common challenges. This would also help to consider how the projects could fit together to define a common toolchain from various independent open source projects (as the Unix philosophy has shown).