Build builders, not projects: what Helpful Engineering learned about innovation and open source

What does it actually take to turn a good idea into something that can be built, trusted, and used in a crisis?

Six years ago, Helpful Engineering did not exist. Then a pandemic hit, a group of strangers found each other online, and within months they had built a movement that would eventually mobilize more than 20,000 volunteer engineers, scientists, and makers around the world.

In this episode of the Modern Industrialist Xchange Podcast, Jason Hehman sits down with Amish Chadha, the organization’s Chief Risk Officer, to talk about what that movement actually built, what it got wrong, and what a shipment of broken oxygen concentrators taught the team about the real cost of an information gap in a supply chain.

This is a conversation about risk, failure, and what it takes to build something people can trust when there is no playbook to follow.

The experts

• Host: Jason Hehman, Industrials Vertical Lead at TXI and founder of the Modern Industrialist Xchange (MIX) community. Jason works at the intersection of technology, operations, and workforce in manufacturing and industrial sectors. Connect with Jason on LinkedIn
• Guest: Amish Chadha, Chief Risk Officer at Helpful Engineering, a nonprofit that organizes volunteer engineers, scientists, and makers around the world to solve real world problems through open source engineering. Amish has been with the organization since its earliest days in March 2020, when it formed in response to pandemic era shortages of PPE and medical equipment. Connect with Amish on LinkedIn

Key takeaways:

1. Most good ideas never make it. That is not a problem to fix. It is the model working.

Amish came into this conversation with a set of numbers most organizations would keep quiet. Helpful Engineering has mobilized roughly 20,000 contributors since 2020, with a core of about 400 people carrying the organization through the height of the pandemic. Together they generated close to 1,000 ideas, of which about 100 became launched projects. Only around 10 of those reached a genuine design and prototype phase. Roughly 5 produced a buildable design. Just 2 spun off into independent nonprofits: Respira Works, a low cost open source ventilator effort that grew from a California garage into a 120 plus person global team on its way to manufacturing in Guatemala, and Project Tetra, which now builds respiratory equipment for firefighting, manufacturing, and other hazardous environments.

Read the usual way, that is a 98 percent failure rate. Amish argues that is the wrong way to read it. Building something safe, certifiable, and manufacturable is one of the hardest things a group of strangers can attempt, and most attempts die somewhere in that process no matter where they start. He described Helpful Engineering as a permissionless invention flywheel, where the failures are the research and development, and the wins repay the whole portfolio.

2. Supply chains do not just break physically. They break informationally.

Amish told a story about a large shipment of ventilators and oxygen concentrators sent from the United States to India during the pandemic. The equipment arrived, but within a week, much of it had gone bad. The lithium components inside the oxygen concentrators had reacted with the humidity during transit and storage, and by the time the machines reached a hospital, they were no longer usable.

The failure was not a lack of equipment. It was a lack of real time visibility into how that equipment would hold up in a different environment, and no shared way for the people shipping it and the people receiving it to catch the problem before it reached a patient. That distinction, between a physical shortage and an informational one, shaped the next initiative Amish described: a framework called SKIS, built with academic partners to notify the parties in a supply agreement in real time when a disruption is about to derail delivery, and to recommend alternatives that still meet the original terms.

3. The advice is not “buy more software.” It is “understand your own data.”

When asked what supply chain leaders should actually take from that work, Amish did not point to a tool. He pointed to a gap most organizations do not realize they have: most companies cannot say what it would cost to rebuild one of their own components in a different region or country, because the data on substitutability was never organized to answer that question. His recommendation was to build what he called a rosetta stone, a way to read legacy SKU and design data consistently across teams, regions, and product generations, and to watch for the point where a design needs so many custom tweaks that it should have been built as a simpler, more interchangeable component from the start.

4. Build builders, not projects.

Near the end of the conversation, Amish drew the distinction that gives this episode its title. Helpful Engineering could measure its output in ventilators, face shields, and prototypes. But he argued the real product was never any single design. It was the thousands of people who discovered, often for the first time, that they were capable of building something at all. In his words, the organization is a permissionless invention flywheel whose failures are the research and development and whose real return is options, talent, and ownership of a network built for the next era of building.

For organizations trying to scale innovation without burning people out or defaulting to hero culture, that shift, from counting finished projects to counting capable builders, is worth sitting with.