[ModernMech]

Frist you'll have to find a researcher you want to work with. Go through all the schools you would apply to and look at the individual faculty pages. Reach out to them and mention your interest with specificity. I get a lot of cold e-mail with a generic "I've looked at your research and it is interesting to me." and those letters go right into the junk folder.

You're most likely to get a reply from a PI who just landed a big grant. They will be flush with cash and in need of students. Departments like to trumpet when their faculty get these big grants, so you can monitor department pages and see who is getting grants. Then you can congratulate them and ask if they need any help, and if so you'd like to join their lab.

I disagree with the sibling comment that you would ever be "locked out" of anything. You might not be getting in to the tippy-top programs as the competition there will be quite fierce, but there's so many programs I think you can find something somewhere. You don't even have to have that much relevant experience in the field. Some programs have catch-up coursework for students who are cross discipline. I did an undergrad in Physics and switched to a Ph.D. in Computer Science, and really didn't know much going in.

So I wouldn't worry about being behind. And honestly, if you've been in industry a while you might have quite a leg up on 22 year-olds from a project and time management standpoint.

[jrumbut]

Having almost finished, it's not just possible it's relatively straightforward and not that uncommon. I didn't even have a good undergrad GPA.

You do need an advisor with some imagination to leverage your skills and fill in gaps that most 22 year olds don't have.