To be perfectly frank, your post comes off more as trolling than as a useful commentary.

The very first question I got from industry was: Is this patented? In other words, can we take this technology and dispense with you?

On the funding side, how exactly am I hoodwinking investors by creating a useful product? Are you assuming that my goal is to sue the world?

I personally think litigation is a fruitless endeavor where the only people who win are the lawyers. I would much rather be in a situation where I develop a useful product than to try and enforce rights. However, the auto industry is filled with very smart people that are literally employed to find a way around letting you make a living off your work. I’m under no illusions that I’m a small fish in a big pond, and I think patents play a useful role in this scenario.

Yes, I would like to have a successful startup. You can call it gold fever or whatever, but I’m still very aware of the harsh reality that most startups fail. There’s no gold in failure. Patents offer a modicum of protection against one of the ways they fail, namely, being crushed by incumbents with greater resources / better name recognition.

I am aware of the public disclosure limits. I’m also aware that Alice v. CLS Bank invalidated the “do it on a computer” class of patents, which as far as I can tell doesn’t apply in my case. I think a normal human would be hard-pressed to figure out the direction and magnitude a ~6 dimensional near chaotic combustion system would move next in less than a millisecond.

Slashdot seems so fundamentally hostile to patents that there is no rational discourse on where they are useful.

No. If they have a valid patent claim, you can try licensing their rights. Or try to find a way around their claims by being innovative. The difficulties are (1) when the patent office grants patents that absurdly obvious (e.g. Amazon’s one click) and (2) when the courts allow absurd cases over simple functions like rangeCheck().

I spent the past four years of my life developing this tool for engine combustion, which exists primarily as software. If I had spent the past four years developing a better pitchfork, would you deny me a right to a patent on that tool? What is the intrinsic difference? They are both tools.

Math is an abstract mental concept that could be applied any number of tools. This is why I agree it should not be patentable.

My work has fundamental differences from those references, but I know for a fact that many of those people have patented their work. Furthermore, patents force people to be public about their work. Would you prefer all modern advancements to be lost to trade secrets known by a select few? That’s a very real alternative when people can’t protect their ideas with patents.

Finally, copyright might be pretty close to forever these days, but patents are trivially short when you think of all the work that needs to happen before you even get a commercial product out the door.

Quite honestly, your reference is nothing like my work.

They have an accelerometer strapped to an engine measuring vibrations to detect if a misfire has occurred, _after_ the fact. There is plenty of prior art for this.

I'm taking microsecond-level pressure measurements inside the cylinder to predict which way near chaotic HCCI combustion will move on the next cycle _before_ it happens. I’m not aware of any prior art on how you can actually predict this behavior a priori in an actual engine experiment (not simulation).

The best work I've seen is from people treating their uncertainties as random variables. This is a far cry from actually predicting what will happen because those random variables have huge sensitivities. I'm also not aware of any random variable approach that has been shown to tolerate engine transients.

Copyright offers zero protection against clean room implementations.

I’m a graduate student who wants to take my thesis research* with machine learning technology for engine combustion control and develop it into a startup company. Being a student, I must publish all of my work (there is no room for trade secrets). There are also many incumbent players in the field of engine combustion control that could easily squash my startup with better brand recognition. How else, besides software patent protection, can I protect my startup company? I honestly don’t see any alternative. * Short video description: https://www.youtube.com/watch?...

I'm using it to run a real-time machine learning algorithm for near chaotic combustion control. The ultimate goal is higher fuel efficiency with less CO2 emissions. Here's a YouTube video about the project: https://www.youtube.com/watch?...