Future proofing

[Dfinney]

I’m looking for advice on considerations when designing a project that will be maintainable for 10 years. Something with digital and analog I/O and built around a microcontroller and touchscreen HMI. How can one ensure that if a hardware component fails in a couple of years that the project will not have to be redesigned and built from scratch? I notice the some hardware from China could be “retired” after 2 years!

[PartialDischarge]

LT had a no obsolence policy, dont know if AD will keep it that way

Slide 6, https://www.slideshare.net/mobile/MichaelSchreiner2/linear-technology-support

[thermistor-guy]

... How can one ensure that if a hardware component fails in a couple of years that the project will not have to be redesigned and built from scratch?...

Short answer: you can't.

When I worked in the Telecom industry, we (my colleagues and I) designed products meant to have a supported life of 30 years. Component choice was always an issue. We would avoid single-source components except in special cases.

Even so, End-of-Life buys were common. When a component was about to go out of production, we'd buy enough to last the expected production and service usage. That would often prevent a redesign, but not always -  redesigns would still happen due to part obsolescence.

In some cases, redesigns would occur before obsolescence became an issue, because the redesign would provide a significant competitive edge (cost reduction + performance improvement).

[MosherIV]

Hi. Welcome to the forum.

Pick components for certain industries can help because some industries demand lifetime support.
Military or automotive sector demand 30 year lifetime, so the component manufacturers will  often sign up to support this, at a price.
Automotive is sufficiently high volumn that the component prices are not extortionate so try looking at automotive grade components.