Pololu Blog » Engage Your Brain »
Rethinking open source in the context of the coronavirus pandemic
We are into week 6 of emergency operations. Our day-to-day routines are largely unchanged from what they were last week (we continue to ship all orders on time with a reduced on-site staff), so please see my earlier posts for more details about that. After more than a month of a new normal setting in, we are striving for a balance of avoiding complacency in daily operations while planning for a future that will likely never be back to how things were a few months ago.
One of our team members passed away (probably not directly related to the coronavirus)
On the complacency front, we were reminded of the stakes when one of our employees unexpectedly passed away at the end of last week from a sudden illness that, from the limited information we have available, was not related to the coronavirus. She was not part of our reduced, on-site staff, so I last saw her in person six weeks ago, and she was in contact a few weeks ago. As the toll of the pandemic mounts, more and more of us are going to be hit increasingly personally, from losing jobs and businesses to missing out on pivotal moments like being present at a child’s birth, to literally life and death experiences made even more painful by new restrictions on being with loved ones and being able to mourn.
However things play out and however bad they get, let’s try to be part of making things better. We can be responsible, supportive, useful, thoughtful, helpful. Many of us are suffering, and probably the only ones not afraid are the ones too unaware to know they should feel some fear. But we can acknowledge the fear without letting it completely overpower us, and we can still look for places in our lives where we can make a difference and decide to make things better.
Rethinking open source in the context of the coronavirus pandemic
The rest of this post is about a longer-term strategy I am thinking about in response to the pandemic: moving toward more open-source projects (both for software and hardware). I would very much appreciate any thoughts and advice people have on the subject.
I wrote about open-source hardware exactly eight years ago this week. I just read it for the first time in maybe five years, and although I feel like I could have written the same thing recently, in many ways I’m a different person than I was then, with the usual progress one would hope for from living another 25% longer, supplemented with extra jolts to my system from things like my baby dying the day before he was born over five years ago and the coronavirus pandemic we have all been shocked by this year. It took me most of those five years to really be able to move on from Dez dying, and when I was locking up Pololu on the Friday a month ago after the first week of escalating government-mandated shutdowns, I really thought I might not be reopening it for weeks and that there would be little chance of Pololu surviving.
Having weathered the past five weeks of emergency operations and being one of the lucky businesses to get temporary funding via the Paycheck Protection Program (PPP loan), my outlook about Pololu making it has substantially improved, and with solvency likely assured for at least a few months, I am thinking about strategic changes for medium and long-term survival in a very different world where the future looks especially uncertain. Here are some of the changes to the world and to Pololu that make open-source projects much more compelling than eight years ago:
Longer-term changes preceding the coronavirus pandemic:
Laser cutting Zumo blades at Pololu on 22 April 2020.
- Pololu’s manufacturing capabilities and sophistication have improved substantially over the past 8 years. (I still believe that open-source hardware makes more business sense for those with manufacturing capability since competitive advantages would move from product design to production capability, which is more difficult to copy.)
- Copies of our products and documentation and other IP have become a big problem. Even in cases where there might not have been outright fraud intended (e.g. counterfeits being falsely sold as products made by Pololu), consumers of knock-offs take up our support staff time, introduce confusion and uncertainty among customers who have genuine versions of our products, and generally dilute the value of our brand. At least for some products, it might be more pragmatic to officially open up the designs than to fight the copies.
- Our brand and processes and relationships (e.g. with our distributors) and other aspects of being a successful business have grown. This gives us value and competence beyond our designs, so opening them up would be less of giving everything away and undermining our ability to remain in business.
Previously-known arguments for open source that have new weight because of the pandemic:
- Given the increased likelihood that many companies will go out of business, having open-source products can give potential customers more confidence in designing our products into their own products or curriculums. Even if Pololu were to go out of business, at least our designs could get produced somewhere else with less effort than having to design in a new product.
- Employees working on the products might be more motivated to work on them since continued existence of their creations would be less dependent on Pololu continuing to make them.
New arguments for open source specifically because of the coronavirus pandemic.
- Pololu was not particularly well set up for remote workers. While we are working on improving the situation, security and similar concerns might add more friction that would continue to make some kinds of remote work/projects impractical. If they were instead completely open-source projects independent of Pololu, security (as far as keeping designs secret) and supporting external connections to Pololu internal systems would be less of an obstacle.
- We have been soliciting donations to help Pololu survive this crisis. Potential donors might be more encouraged to donate if they had some confidence that they are supporting existence of products and designs that will continue being available even if Pololu were to stop manufacturing them.
- There are now many more people out of work (whether paid or not) around the world who might be open to donating their time and expertise to help on Pololu projects.
- There are now many more students at home or recent graduates without jobs or internships that could especially benefit from looking at or contributing to our designs.
- The pandemic has increased awareness about potential benefits of manufacturing locally, or at least being less dependent on one region or country. Perhaps that will alleviate some of the price pressure that makes it more difficult to open up designs (if you have to compete with the cheapest place in the world to make something, it’s harder to just give them your design).
In short, the heightened uncertainty about business collapses, shortage of money, and physical separation/decentralization that the coronavirus crisis is forcing on us all substantially tip the balance in favor of moving toward more open source projects in organizations like Pololu.
I would appreciate any advice or thoughts any of you have on the topic. Here a few areas you might be able to comment on:
- What do you think in general?
- I am considering specific product-based fundraising campaigns in which we could open up some existing products after exceeding some donation threshold; how does that sound?
- Are there some existing open-source projects that you would like Pololu to start contributing to or manufacturing?
- What’s the latest on open source hardware, in terms of standard licenses, business models, examples, etc? (I generally feel like I hear about open source hardware less than I did 8 years ago, and that there have been some notable disappointments/sell-outs that have dampened the movement, but perhaps I just have not been paying as much attention.)
- What are the best open-source software tools for electronic CAD and mechanical CAD?
- Are there notable recent success stories in terms of open-source physical products?
- Are there any notable examples of companies or organizations attempting to produce open-source physical products?
Thank you for your continued support, everybody! We are working hard to be worthy of it and to do our part to make things better.
Re: "it might be more pragmatic to officially open up the designs than to fight the copies." -- Wouldn't dealing with
support requests for counterfeits still take up staff time?
Re: "The pandemic has increased awareness about potential benefits of manufacturing locally" -- I'm not sure this is true. It might show the opposite, that global trade is so important, especially during a pandemic, that government should be prevented from interfering in it, via tariffs, for example. At least, I hope so.
Re: "we could open up some existing products after exceeding some donation threshold" -- it might be more sensible to solicit donations after a particular product is opened up.
And finally, you didn't mention it, but your various bits of firmware could be made into free software.
I hope you are doing well, and thanks for your comments.
Yes, dealing with support for units from other sources would still take up our time, but I'm hoping that the balance would get better. As in, initially, keeping things closed might help prevent this problem. But once there are enough copies out there, anyway, we're getting the negative of extra support overhead without the positive of others possibly supporting our customers (or the more resourceful ones supporting themselves better, e.g. with questions like, which resistor is R3 in the schematic?). So, I just think the balance might be better for us now than it was eight years ago.
Surely there are some benefits to manufacturing locally, and surely some people are more aware of them now. That does not mean global trade is not important or that it needs to be resisted via government intervention. I think there might be more people willing to voluntarily buy something from me for $4 that they might have ordered from China for $3. If I'm happy making that thing for $4, isn't that overall a good thing?
Why would it be more sensible to solicit donations after a product is open?
And on your last point, I'm not sure if you are making a distinction for "free software" vs. open source. The scope I am thinking about includes software (including firmware), electronics, and mechanical designs, and as far as details of licenses or implementation, I'm probably thinking about the equivalent of making the designs/projects/whatever public domain.
Wonderful post, thanks for sharing and detailing your businesses' experience through this crisis. First just want to thank you and your team for everything you've done to support the Open Source Hardware community already. I worked at LulzBot for nearly five years (former President of the company), we were proud to carry your products and our customers loved them!
Agree with many of the points that you raise throughout. In particular, under "New Arguments", your points about enabling local manufacturing and future availability (beyond the company's time in business) are especially resonant right now. We're seeing that in-practice here and I think it's an important point for the future. Reminds me of work by RepRap, Appropedia, Open Source Ecology, and others.
For creating with open source tool chains, the most popular (from my vantage point) certainly seem to be KiCAD for electronics design and FreeCAD for mechanical design. Both are active and rapidly improving. There are many examples of projects I could see Pololu contributing to and having a strong value-add, but that is also a function of your overall strategic goals for the business.
In terms of the state of the community, I'd strongly encourage you to consider the work we've done at OSHdata. We published a 2020 report reviewing over 400 certified products, including their licenses, country of origin, product category, and more. Overall, the program has been growing rapidly and since we published the report, the list of certified projects has basically doubled in 2020. The community is alive and well! There are many companies to consider that report as examples. More than I can fit in a comment here! I hope you find it useful, you can find the report here (it's free and shared under an open source license).
The one perspective I'd encourage you to consider is that hardware is the critical entry-point, it's not all. There are companies on that list going to market with three-prongs of hardware, software, and services, For bootstrapped companies like Pololu there are many benefits to taking this approach. It also enables moving into higher price points and serving enterprise and government buyers. This can be a nice logical extension of a business model, beyond the lower average sale price products (<$100) typically offered by companies focused on components and electronics. Beyond 3D printers, there are examples of Open Source Hardware robots, high performance computers, and more that can fetch over $50,000 a piece and create tremendous value for the end users.
I would welcome the opportunity to connect with you and your team to discuss this further. You can contact me with the email I used here for my Pololu account, on Twitter @harriskenny, or by emailing email@example.com.
Either way, thanks so much for everything you do. Hang in there, the world needs companies like Pololu!
There is an alternative to straight open-source hardware that provides longevity benefits. And that is putting the design into escrow along with a promise to releases it when you stop manufacturing it. My software company has escrow agreements with several of our customers.
Curious (if I may constructively ask)... Can you share a salient example of a product of yours that gets notably knocked off? Why do folks asking for support think it came from Pololu, does your logo get copied too?
If you go to eBay or Amazon and search for "Pololu stepper driver", pretty much all the results are knock-offs.
For another example, search for VNH5019. You'll see multiple examples of boards that have exactly our parts arrangement and label layout (i.e. it's not just about a functional solution to a problem).
1) Countless clones, many of them low-quality, that swamped our customer support
2) Very little "giving back" by the cloners in the form of open sourcing improved designs
3) "Race to the bottom" of pricing, making it uneconomic to compete with our own cloners
So we eventually gave up on OSH.
We did, however, double down on open source *software*, which was much more successful in generating community contributions and participation.
Basically, I think there was a brief window in the Arduino era of the late 2000s when hardware was simple enough that lots of people could spin and remix open source designs. But now in the era of ghz Linux SBCs and BGA chips, regular people can't usefully contribute to the hardware and it's mostly Chinese hw cloners that benefit.
A few companies are still doing with with OSH, such as Adafruit. But they're the exception, not the rule.
And I think that tenuous finances are not a good reason to OSH ("if we go bust, you'll still have the designs"); if anything the price competition from cloners will make that more tenuous.
Thanks for sharing your experience and perspective. It's good to keep in mind that I probably hear about open-source hardware more from the few people promoting it than from people like you who have actually gone out and tried it on a big scale. It sounds like you found a balance, though (on the software side), and I wonder if there still might be (profitable) room for some OSHW products at Pololu. One advantage we have is that we have a broad array of products, so we can try it with a few of them to see how it goes.
I am not an expert in electronics; this is just a hobby for me. But being an amateur never stopped anybody from sharing his thoughts online...
First, I definitely accept to pay more for a product that is open source and manufactured by the original product designers. So feel free to moderately increase the price of any newly open sourced product. One reason for this is that it feels fair to compensate your company for added support costs, extra documentation costs, and more competition. A second big reason is that I learn a lot by looking at the internal designs of the hardware and software. Not that I could improve it or fully understand it, but it enhances the educational aspect of the hobby.
I don't know if this makes any sense but for me it would be ok if you were to release the product designs with a non-open hardware license that does not allow competitors to copy the design as long as you sell the product, but that allows customers to know everything about it.
I like a lot the idea of product-based fundraising to "free" popular products. You could offer to free the product after certain number of units have been sold (i.e, when you have recovered your product design expenses). Or with every purchase you could ask for a contribution to free the design, kind of like a ransom. In any case, I would be happy to see a product I love become open source and "live forever" online.
One final thought. I wonder if the push for “Right to Repair” laws could reduce the gap between open source and closed hardware projects. It is too unfair for customers when a company fails and all the expensive gadgets become unusable and unfixable.
Best regards to everybody!