Dassault Systèmes’ annual Science in The Age of Experience Conference has become one of my favorite conferences. It is a platform for thoughtful and innovative discussion centered around how science Continue Reading
Dassault Systèmes’ annual Science in The Age of Experience Conference has become one of my favorite conferences. It is a platform for thoughtful and innovative discussion centered around how science affects so many aspects of our lives, as well as an intermingling of distinguished scientists, engineers, and other professionals from all walks of life and disciplines. Unfortunately, due to the COVID-19 pandemic this year the conference could only take place virtually. However, I was very fortunate to have had the opportunity to grab some time with one of the main keynote speakers thanks to the Dassault team who have partnered with Prof. Gershenfeld on many impactful projects.
Dr. Neil Gershenfeld, Director of MIT’s Center for Bits and Atoms, meshes together the physical and virtual worlds through quantum computing and digital fabrication of the IoT, and pushes the boundaries of what can be done in order to create a more sustainable future. He has an impressive list of accomplishments and credentials behind him that are well-highlighted by the IAAC (link here), with research advancements including:
“one of the first complete quantum computations, using nuclear spins in molecules; microfluidic bubble logic, with bits that transport materials as well as information; physical one-way cryptographic functions, implemented by mesoscopic light scattering; noise-locked loops that entrain on codes, which led to analog logic integrated circuits that use continuous device dynamics to solve digital problems; asynchronous logic automata to align hardware with software; Internet 0 for interdevice internetworking; microslot probes for ultra-small-sample structural studies; integrated 6-axis inertial measurement based on the dynamics of trapped particles; charge source tomography for electric field imaging and intrabody signaling; and additive assembly of functional digital materials.”
In our conversation together, Prof. Gershenfeld talked about his network of Fab Labs (fabrication laboratories) that are multiplying via Moore’s Law. Moore’s law is not a law of physics, but rather a law based on empirical observation of production rates predicting that transistors on a chip will double every two years. In other words, these labs are multiplying very quickly around the world.
More importantly, these Fab Labs are encouraging and enabling non-traditional collaborations and innovations spanning across disciplines, cultures, and socioeconomic statuses – all via the language of technology. The partnership between the Center for Bits and Atoms (CBA) with Dassault Systèmes has really facilitated the spirit of all things DIY and made it possible. To learn how to use the tools available at a Fab Lab, Prof. Gershenfeld teaches one of the most popular classes at MIT called, “How to Make Almost Anything.” We talked about some of the impacts this work has had in medicine, and globally in improving the accessibility to science and innovation.
Alice Ferng, Medgadget: Tell me more about the Center for Bits and Atoms and what you do. How do you continue pushing limits and where do you see things headed in the foreseeable future?
Dr. Neil Gershenfeld, Director of MIT’s Center for Bits and Atoms: As background: we were part of doing the first quantum computations and a part of creating what’s now called the Internet of Things. One of my students built and runs all the computers at Facebook and one used to run all the computers at Twitter. Those are some of the pirate things CBA has done looking at the boundary of digital and physical.
We were part of creating the minimal synthetic organism- creating life from scratch. And, broadly we’re looking at how bits meet atoms, and that’s at many different length scales and timescales. And then around that, we have a very close collaboration across that whole roadmap from very short-term impact to very long-term research roadmaps with Dassault Systèmes. One way you can understand it is digital computing went through Moore’s Law era. We’re doing something similar to Moore’s law, but going from literally bits to atoms: from democratizing access today and immediately fighting the pandemic, all the way to research roadmaps to the replicator.
Medgadget: What inspired you to get into this field of work?
Dr. Gershenfeld: I was never not in it. I was always interested in the boundary of digital and physical. You can trace it back to when I fought because I wanted to go to a vocational school where you would learn to weld and fix cars, and I was told, no, you’re smart. You need to sit in the room. When I was at Bell Labs I had grievances from the union because I tried to go in the workshop and they would say, no, you’re smart enough to tell somebody what else to do. Eventually you came to understand that it’s essentially a mistake made in the Renaissance, which is when the liberal and the illiberal arts diverged, and what we’re providing is new means of expression that involve literacies of digital design and digital fabrication.
So I’ve been doing this all my life. It was only when I really got the chance to create the Center for Bits and Atoms, which grew out of an NSF investment to make and unusual facility where we can make almost anything on any size – from molecular size to building size, and then also build the research community around that. And then with that some of our output is basic research. Some is broader impacts things: we started a network of thousands of community Fab Labs. And then there’s a core group of companies we work with who are much more than research sponsors – the real collaborator through the arc of it. So we are just selling the whole arc of the tools that make this possible. Then jointly with many of the companies, we both work with places like Airbus on what does it mean to make jumbo jets without traditional global supply chains, and that can change their shape and Toyota on cars and old Oldendorff on ships, things like that. And so we intentionally manage this portfolio of basic research mixed with short term applications in ways that they there’s very strong synergies across them.
Medgadget: Tell me more about the synthetic biology work that you’ve done, and how technologies fit into this picture.
Dr. Gershenfeld: We’ve been working with JCVI (J. Craig Venter Institute) and a few other collaborators. Our contribution to that is in a few different areas, but in that one, our contribution was the microfluidics for genome transplantation that there’s brilliant work on whole genome design, but then biologists would just pipette tubes of colorless liquid. And it turns out that geometry is really important in biology. A while back, we developed techniques to do micro-machining for microfluidics, and we introduced things like digital logic and microfluidics. And based on that work we developed microfluidic structures to facilitate genome transplantation and image it. And so that was part of the work for the minimum synthetic cell. There’s a paper we have under review right now on implications of that for understanding how minimal genes relate to properties like cell size and function. And then some of the other projects you’ve done neighboring to that are merging machine learning and Big Data with molecular biology. We were able to significantly expand the parts of the genome that CRISPR can get access to. Another one of our projects developed soluble surfactants that let you work with membrane proteins out of cells. You can use these all as digital design problems, but in this case with molecular biology.
Medgadget: What do you see the applications to medicine being with that body of work? How have you been able to scale these efforts across communities and cultures?
Dr. Gershenfeld: Well for applications to medicine, what we’re talking about now is much, much further out. To start with applications to medicine, much closer in is the immediate pandemic response. Why don’t we start with that and then look further out at the longer research roadmap. In this short term, given our unusual feed of tools, we were approached early in the pandemic for ways to help. And I put together a working group, that’s run into a few hundred people from including government regulators, healthcare providers, basic researchers, community activists, startups, and industrialists all looking at how rapid prototyping can help with the pandemic. A number of things have come out of that. Early on, it helped identify that the focus on ventilators was misplaced and what was needed was intervention long before the ventilator.
We worked with Simulia to do computational fluid dynamics modeling of particle flow with a fairly advanced simulation that showed that the widely produced face shields were failing to cover routes where, where particles would get into people. And it led to a very interesting map called a curved crease folding to make curved shields to provide much better sealing. That then led to creating isolation boxes to help provide protect frontline workers, very cumbersome designs were being developed for that, but there’s the beautiful map of the curved crease folding that let’s you quickly make a light, strong structure. Other projects like nanoparticle imaging to characterize emerging filter media were all things that came out of that. And then it’s progressed to projects like instrumentation for longer-term monitoring. There’s fairly basic research to do the work I’m describing. But then, we were working with people all around the world – in particular in underserved communities.
Medgadget: I’ve been involved with multiple COVID-19 efforts myself, including helping lead Microsoft’s PPE efforts with a handful of other amazing individuals – I’d love to hear more about how folks around the world have been able to benefit from access to Fab Labs and Fab Lab equipment during this pandemic.
Dr. Gershenfeld: One of the most interesting things to come out from this is: there was traditional industry that could scale the response slowly and respond slowly. Then there was the do it yourself responses that lacked guidance and could respond quickly but couldn’t scale. And what emerged were distributed production networks, where the research collaboration could help curate best practices. And then you could essentially ship data and produce locally. So colleagues in India produced millions of urgently needed items of PPE by doing this coordinated, distributed production. This is a really interesting moment of collab – bringing together basic lab research and simulation tools with urgently needed healthcare interventions, but not going through traditional routes of production and distribution. All of that is a very immediate impact of all of this.
There’s been a number of efforts like this in collaboration with Dassault Systèmes, where there were lots of efforts to quickly make things like PPE. There were a few things we found – for one, many of them were poorly informed about best practice and there’s real research you needed to do to support them. So things like the colleagues at Dassault, the 3DEXPERIENCE platform, and the lab tools that CBA had helped us curate best practices. And then in propagating them there was this middle ground in between small-scale do it yourself and traditional industry of curating. Again, shipping digitally and producing locally. That’s been something envisioned in the future, but this really drove that in the short term. And it really led to very unusual collaborations: people who had never worked together before thought they would benefit from doing that.
Early on, they were making things like ventilators, or making pulse oximeters, EEGs, and EKGs. There’s a range of currently expensive and hard to get access to advanced medical instrumentation that you can produce with the tools in the Fab Lab. And there’s all sorts of questions about safety and oversight, but you can radically expand access to tools for healthcare. And then there’s a really interesting recursion happening.
Maybe one of the strongest messages from this is that medicine has been based on regulation and rules. Whereas, the progress in software and hardware is democratizing the means, which means that it can be more responsive, inclusive, and accessible, though that also means you need to switch from kind of command and control and regulation to this sort of world of soft power and opting in, which is similar to again, how things like the internet evolved.
Medgadget: What do you think were the biggest challenges with getting the local fabrication processes started? How have you been able to keep these efforts propagating and keep them going?
Dr. Gershenfeld: We had advanced prototyping tools at CBA that we weren’t using that we sent off to community groups and we had emergency room doctors coming in to use our tools. We had a community of groups all across India following computational fluid dynamics from Simulia. Of all of these examples, the hardest thing was there was this fog of confusion and putting together best practices and bringing all these groups together and realizing we needed to all work together. You know, each was started in isolation and, you know, realizing that we needed to work together and we could work together over and over with the supply. It’s just recognizing the collaboration. So then starting to look forward from that, one of the interesting medical implementations is to step back computing scale from mainframes to minicomputers, to hobbyist computers, to personal computers, to the Internet of Things.
In the same way, manufacturing is scaling from traditional factories to a network of thousands of communities of Fab Labs. And then we’re working on rapid prototyping of rapid prototyping machines and maker machines. And then further upstream in the research, the most fundamental things we’re doing are going from traditional manufacturing to assemblers and self-assemblers. And in a very literal way you can do that. We were looking to create life and in living systems taking the same sort of principles of how all of life is based on 20 amino acids and doing that for the rest of technology. We’re working with the National Institute of Standards, Materials, Genome Initiative that was misnamed because it didn’t really have the genome and actually really developing the genome from materials. But now again, to come back to medical implications, what’s emerging is that with the tools in the Fab Lab today that are about two times a hundred thousand dollars, you can make much of medical instrumentation.
We started these Fab Labs as outreach for NSF, and they went viral. They’ve been doubling every year and a half. At MIT, I teach one of the most oversubscribed classes called “How to Make Almost Anything” to learn how to use these tools. I started for the Fab Lab, a global version called the Fab Academy. And it’s a very different educational model: instead of being centralized for online, it’s distributed in students of peers, in work groups with mentors and machines locally, and then we connect them globally with video and content sharing. And then my colleague, George Church – he’s maybe the greatest geneticist – was struck by that. With him, we started what’s called the Bio Academy to teach biotechnology in the same way (How to Grow (almost) Anything). We have this really interesting recursion: with the tools in a Fab Lab, you can roughly make a bio lab. You can make PCR and centrifuges. And one of my former students developed the $1 microscope and the CD spinner centrifuge. And so you can make the tools for biotechnology. And so we’re looking towards the future where the Fab Labs can not only make Fab Labs, but they can make bio labs, and then you can develop biotechnology in them. There’s a really interesting boundary to that of things like open reagents and bringing open source projects to molecular biology. That raises all sorts of questions about, you know, safety, efficacy, regulation, but also a great opportunity for open source software. It’s why, you know, Microsoft is almost a completely an open source company and something that it’s really embraced. And what we see beginning to happen is that emerging in molecular biology.
Medgadget: How do you fund the labs? I assume it’s basically the same for each Fab Lab such that each one can be outfitted with similar machinery and tooling?
Dr. Gershenfeld: So it’s like the growth of the internet in that the first 10 I paid for. But now there’s thousands and they’re doubling every year and a half. It’s a mix of a for-profit, non-profit; a range of funding. And again, one of the most interesting kinds of funding is: we’ve been working with companies like Dassault Systemès on that as part of the collaboration. We’ve had a really fun sequence where every year with Dassault. So out of the thousands of labs, CBA can only do a few of that so we spun up a foundation to help manage this network. But every year we pick one really interesting place in the world with Dassault Systemès, and jointly deploy one of these labs where they support it and we provide a team.
We did one in Rwanda. We’ve done one in Bhutan. One in the Southern most city of the Earth at the bottom tip of Chile. One coming into Nepal humanitarian relief. And again, there are really interesting rippling implications – you may know that Bhutan is not based on GDP (gross domestic product), but ‘Gross National Happiness’ (GNH). It doesn’t mean they’re happy, but it means they measure well-being and not money.
What we did there with Dassault Systèmes was so successful that they want to create a national network. In order to do that, we’re deploying million-dollar Super Fab Labs with the tools to make their own Fab Labs. We did the first of that in the South of India and Kerala. The next one is coming in Bhutan. That’s in turn related to there’s a Fab City Initiative that started in Barcelona – Barcelona has great design sense, but 50% youth unemployment. So much more than being a smart city, one of my colleagues became the city architect and is deploying these tools as part of urban infrastructure so that if you live in the city, you have the means to produce. That’s all leading to a Fab City Initiative with right now about 30 leading cities around the world.
Another piece of this is legislation in the U.S. House and Senate for universal access to digital fabrication as a new kind of “right.” And so all of that is like the growth of the internet, where initially was funded as a research project. Now there’s a range of types of funding, but you maintain compatibility across it.
Medgadget: That’s truly amazing and inspiring. Has it been difficult for certain locations to source the materials required to continue to produce and replicate?
Dr. Gershenfeld: Yeah, so what we find is there’s a small set of high-tech consumables that right now we need to source globally like precision end mills or microprocessors, but much of the rest can be sourced locally. Upstream research in the lab is: how you make things like make a microprocessor in one of these labs, and that’s something we’re working towards. But one of the interesting collaborations is with NIST, the Department of Energy, and a group called Curium, which is doing open materials and looking at how to source natural materials to produce the range of advanced materials for technology. There’s also really interesting research on how to produce bio-plastics or biocomposites to expand what you can source locally.
Medgadget: Have you had challenges with literacy rates and things like that in terms of scaling out?
Dr. Gershenfeld: The problem is almost the opposite. There’s language – which has not been seriously rate limiting – just because much of this is done in the language of technology. At these community labs we’ve been developing, we’ve had great sessions with the Dassault team and what we find consistently is exactly the same profile of brains and the people I work with at MIT anywhere we do this – the world’s full of great people, but the traditional pathways for them are vastly limited. MIT is thousands of people and the world is billions of people. And so we find these bright inventive people anywhere we do this in the world. What’s been missing is the means for them to work that we’re describing, and then ways to mentor them. And so what’s been hard isn’t the literacy, or finding them – it’s just building the capacity to do the scaling of it. I’d say the technical roadmap we’re describing, I thought was hard. That’s actually going pretty well. What’s been harder is the kind of organizational innovation to build the capacity to scale with this world.
There’s this really interesting interplay. We spend billions of dollars in new businesses, and there’s a lot of mid-sized businesses that are going to fail in this world in the same way that the minicomputer industry died when personal computers came out. Part of what is so interesting about us working with groups like Dassault Systèmes is that we’re over a 50 year arc, and not short term. We do short term projects, what I mean is that it’s really about thinking in a 50 year arc: what is the future of the planet? There’s been a number of these really thoughtful, interesting companies that in a deep way, think about how to help us live in this new world, you know, much more than just the traditional narrow notion of the relationship between a company and a university and society.
Prof. Gershenfeld’s recent book (includes Research Roadmap)