[ducktective]

Oh I've got many questions :)

1- Isn't a HW startup really hard compared to software? Like the ideas are like yet-another-social-media-site or a SASS but in hw one needs to do the embedded programming, the mechanical design, the software, marketing and *then* the novel idea that sells. How do you do it?

2- How do you manage to find the expertise to build a viable product? How did you find investors?

3- Any tips/books you'd recommend for hardware/engineering related startups?

[jjsylvestre]

1. There are parts that are certainly harder and there is one part I can think of that may be easier.

- Harder: Longer and more expensive iteration cycles, MVP can be expensive, manufacturing and production required to scale, expensive certification

- Easier: Raising money (sometimes) especially if you are a moonshot with a big vision. Good example is Boom Supersonic.

Good news is there has never been a better time to start a hardware company than today. Iteration cycles are becoming shorter due to advances in rapid prototyping and there is a lot of capital available, especially in electric vehicles and sustainable tech.

Peter Thiel talks a lot about this in Zero to One, but much of the innovation that’s been done in the past decades has been in the digital space. We have so many problems that require innovative hardware solutions and I think now we are just beginning to scratch the surface.

2. We all met through Formula SAE in college. This is a great place to meet super talented engineers and is why Tesla, SpaceX, and the other top companies in the world recruit heavily from these programs. It teaches you both the hard skills and the soft skills. If you are still in college, I would recommend getting involved in teams like this.

3. I haven't really read anything hardware-startup specific, but I love Zero to One and find myself rereading it all the time.

Would love to hear other peoples thoughts on this as well. Good questions

[throwawayboise]

I'm not in the field, but from my perspective, aviation is super, super conservative. That is one of the reasons why we have GA still using leaded gasoline in engines that are largely unchanged from the 1950s. F1 is basically the exact opposite, they are pushing the bleeding edge of technology all the time.

[HeyLaughingBoy]

My perspective might be biased, since for the 30-odd years I've been in industry, I have done mostly embedded systems.

Those things you mention are skillsets: software, mechanical engineering, electrical engineering, etc. Sometimes you find them all in the same person, sometimes you have separate bodies doing each one, but at the end of the day it's a staffing issue.

I'd say it's less that it's "hard" and more that iterations are slower and costly and the less capability you have in-house, the slower it is. If you have a full machine shop and a Stratasys 3D printer onsite, a lot of things can happen faster. If you're doing garden-variety industrial automation, there's far less risk than designing state of the art humanoid robots.

There is a spectrum of difficulty, like everything. In my case, I was doing this stuff freelance, in my spare bedroom at age 25, so it's not that hard.