Ask HN: Any small electronics/hardware manufacturers need a custom PLM? | Hacker News Reader

Ask HN: Any small electronics/hardware manufacturers need a custom PLM?

3 pointsall22 comments

I'm looking at building a custom solution for a manufacturer in familiar with. Their current processes are "on paper" but digital. IE, share drives and PDFs. Documentation is scattered, old stuff is a mystery (unless you know where to look on the share drives), and most pressing to me is the bills of materials for end item products are split across three different sources of truth; one for engineering, one for production, and one for purchasing. Second on my list is locating documentation and files associated with a particular part number.

What I'm curious about is your pain points at small manufacturing companies. Have you faced similar issues? What solutions did you land on? What worked? What didn't work?

Would you find a somewhat opinionated but configurable PLM with a basic feature set useful? On-prem (for cmmc and security requirements)?

View on HN

    …unless you know where to look on the share drives…

A place where I worked had almost all actively used manufacturing files stored in a network folder labeled "do not use".

The MRP software they used had so many layers built on top of old layers that it would occasionally throw COBOL errors. This was a 3rd party system too, not something cobbled together in house, but no one had bothered shopping around for something better. That was mostly because the accounting department used it to, and while to most manufacturing departments a 20-year-old computer is as useful as a brand new one, you're lucky if you get an accounting department that knows how to let a spreadsheet do the math, instead of manually typing it into a calculator.

Regarding your questions, what you've come across is sadly pretty common. The best advice I can give is that for anything to be accepted it absolutely must be stable. It can also be helpful to offer multiple options for each change, let the employees try them out and pick what they like, then build on from there, working your way to a full implementation of the new system. If, after several iterations, you realize that you need to change something you had implemented earlier, it's okay to do so occasionally, but the advantages over the old method must be clear to everyone using it.

Ironically it is not a problem that can be solved by yet another software system to learn and use properly.

The mess you and many others have found is symptomatic of a lack of cohesion, standards and conventions. It is just too hard to retrospectively sort out such a mess. When everybody is hard at work on current projects, where is the time going to come from to clean up the past mess?

The only feasible solution is to implement a clean new system for all work from some point in time. Of course, it requires management fortitude and staff buy-in -- neither being easy but absolutely essential.

The most viable approach for you is to do an absolutely stellar job with your current client. Then using them as a reference site approach comparable firms with your proposal to fix their pain point. That is where your configurable PLM comes into play. But you will need to do all the work. Clients will not buy such software and then invest in yet more time and effort (i.e. money) to learn it well enough to configure it for their needs.

Reply on news.ycombinator.com