Tyler Mincey: Hardware is Hard, so Is Software: Product Development in Bits and Atoms

Tyler Mincey (@fictivetyler) was speaking at Smashing Conference about hardware and software design, these are my notes from his talk.

Strong design works in deep partnership with strong engineering.

What is a "product"?  Beautiful sturdy products, not first prototypes. 

Hardware

Typical team has designers, engineers, product mangers, test, specialists, procurement.

Concept to mass production takes roughly 10 months.  Mass production is about 200k units per day.

Start.  Concept review - high level goal, features, design for the product.

8+ weeks.  Project proposal - 90% complete design.  Tool release, confidence in industrial design and mechanical reliability.

9 weeks.  Protorype - first functional build.  Test functionality and reliability.

6 weeks. Engineering validation test.

6 weeks.  Design validation test.  Minor changes and safety testing.

Mass production validation test, first final products.

• R&D
• Build
• Prove
• Scale
• Post ramp

Crowd funding often kicks off at the prototype  stage.  Offer rewards at EVT stage.  Mass production fundraising.

Sweat the small stuff, root cause of all issues in a prototype.  1% issue at manufacturing scale is a real problem.  This takes a lot of effort, you need the space to do this.

By moving slower people got a chance to build a relationship with a device, don't get that with the quicksand of continuously deployed and improved UIs.  Encourage delight not expectation.

Everyone on your team is a designer.  Any time you are making a decision within constraints is a design decision.  Innovation happens by moving constraints, not just designing within them so you land in a different design space.  Empower everyone to push the boundaries.

Software

Software is very similar to hardware.  Follows the same process as above:

• Concept review
• Project proposal
• Iterative build process (four three week chunks)
• Launch
• Fundraise + transition.  Support.
• Transition to sustaining team. 

Lesssions

Share and collaborate - hardware has a lot to learn from open source, the open source hardware movement is only jut getting started.

Ship early, ship often.  Hardware needs innovations in tooling for this to be practical.

Build social, build community - helps defend from fakes and competition.

Iteration and schedule are key.

Improvements happen through fundamental investments in the platform, not the bells and whistles. 

See Dieter Ram's principles of good design.  No 6:  Good design is honest.