Obviously you are new to this. The first step to anything is coming up with a name. A name that can rally developers to your cause. Act fast, all the good names are taken. Without a name, you can't even market or sell or profiteer from your OS - and what good is that. ( Read Dilbert for details. )

The problem you have is pretty much it has all been done. Who wants to spend their time on re-creation rather than innovation. So you need a hook, something that is a minimally viable kernel.

So what is the thing that this new OS is to do better?

Back to Dilbert, you need a project you can market to developers as something worth working on. THAT is how you get other people to code your operating system. It also helps to have done some of the heavy lifting. That minimally viable kernel, that shows you are more than a dilettante or a provocateur.

So what is your pitch? Why do we need a new operating system, why do we need to code it for you? Go....

( I was being "jokey" in the 1st sentence. )

"No, help ME write MY operating system!" :-)

Some aspects to consider are: Topic I: Architecture of operating systems and process management. - Architecture of Operating Systems. -Process process or threads. -Synchronization and communication between processes. -Planner Planning -Strategies of interlocks. Topic 2: Memory management -Generation of the Main Memory. -Mechanism of Virtual Memory. Topic 3: Management of entry / exit. -Software of I / O. -System of files.

you must first learn Architecture of Operating Systems. And understanding the computer Hardware. And the best step to read the code of open source OS.