I would like it to be more plannable and sharable on a corp level.

Maybe something like you have 100 arbiter assault kernels and want to research those. Instead of giving you random knowledge (like atm) you get to choose which kind of equipment you want to research. Lets assume arbiter assault kernels are worth 100 points of research and you only gain those 100 points on the selected field.

The current system is very very unfriendly to new(er) users as there is no way around it.
Maybe something like the possibility to take 10 kernels + a t3 item to "invent" a t4 item (or ct) at a given chance % would make higher tech available to users that didn´t grind 10000 kernels.

This would also be a workaround to the current prototypers belonging to the big corps that dont sell anything of their tech on the open market, which would also mean an increase of high-tech items available on the player driven market.

Personaly I would prefer to be able to semi specialize if not specialize.

What I would like is for tech to be linked closer to the color and lvl of the kernel you eat. IE if i eat only green kernels I will only recieve tech for their modules. Also higher lvl kernels would give you progression on higher teir item, and lower tier items at the same time. IE eating a green obs kernel gives you a small amount of progression on a t4 item, more on the t3 etc etc.

Something like this could let you semi specialize while not being able to get too specialized

Research is all or nothing, having 90% of T4 completed has no benifits. Being able to choose an area, say engineering, to complete befor starting other areas would be good.

Edit: i wrote BS
Leave the current system how it is. You can provoke new wave of unsubbing and flaming.

In my opinion players created many dedicated accounts for prototyping. If you want this prototypers carebear for research one more time: this prototyper will be useless. Why? Because whole corp worked for few dedicated accounts, and worked really much. And now you want to put corporations into situation when they must work for research what will erase itself after some time/ct creation?

All fine with curent system: after prototyper gained knowlege he need huge ammount of mats to make this prototype. All research depends of kernels you using, and people fine with it: want a best tech - do farming. Gained desired tech: now you can sink mats to make proto.

Now its too late to make a brand new experiment with core functionality. This changes should happened during beta-test, not after game release.

Im highly sure that current prototypers will drop playing if devs will implement sort of "erasing of KB". Prototypers now had a problem to get a mats for prototypes, and you proposing to add new butthurt for kernels.

New "experiments": new unsubbing wave.

At the moment marked are not filled by prototype orders. Its very important moment what should be attached to this discussion. Ability to sell or buy a prototype from marked is tying prototype to the NIC currency. Given that the market has not developed yet i can make a conclusion that noone now can realize what changes should be made for research system.
Also any change, what was proposed will made nothing to kernel flow.

Static knowledge base (research will be allways with you, no matter of its usage), it is suiting this game perfectly: Average New corporation should spend half of year to gain full tech tree. If game will be expanded, this time will be increased. After new tech researched, corporation needs *** of resources to create high tiered items. Researched tech tree is a hole for a big nic sink. Randomized research good for nic sink because you need more kernels. Unlimited KB size is a good start for huge nic sink. Linearity of research is a nic sink too: you work hard for 100% result. And then this result will cause a nic sink.
What can be changed?
Add specialization? You can specialize now, researching different colors of kernels.
Add specialization to types of equipment? Why? If you can research desired equipment with a bunch of kernels (nic sink). Maybe its better to change research speed
Change observer's kernels mechanics? Why? Observer kernels designed to finalize t4 research, what needs huge amount of normal kernels. Those corp who can farm observers in enough quantity can boost the hardest t3->t4 research, after a hard work to equip your warriors. Risk-vs-reward.
Make "dynamic" kb with research what erase itself after usage (like prototypes)? It will change nothing, it only shift nic sink to other branch of game. Kernels will be allways on a market, you can see it even now. No matter how many dedicated prototypers are in corp.

Good nic sink = more time online = larger game.

Ok here is my view on the issue. During the last months, many changes were made that did indeed change the kernel flow. The droprate was reduced some time ago and with the recent changes to NPC-Spawns farming the required kernels is a lot harder or nearly impossible(High-end indu kernels for example). This means a new Corp right now has to farm about 10 times more(this is a wild guess) to be on the same tech-level as someone who farmed a few months ago. This is a very huge and dare i say unfair advantage for veteran players and Corps.

Also this system is not well balanced as it promotes large corps. They will get the farming done faster while small corps may never be able to get all the kernels. Also if a prototyper stops to play it would not take as long for a large corp to replace him...maybe they even have stockpiles of kernels.

The system itself also discourages kernel trade as it is always a good idea for a corp to have them in stock.

I think a solution could be to improve reverse engineering. You would have a choice to either do it the traditional way, so build a prototype end RE it. The new way would be to just take the item you want a CT of, add the decoder and after that some fitting kernels(or something else...robot parts, minerals you name it). The more kernels you use the better the CT will be. In the long run, the prototype method would be cheaper, but if you only want to create some item for yourself the extra costs for the kernels might not be as bad as buying the item from someone who just made 300% profit because he is the only one selling it.

This way new players can directly start to manufacture high-end items if they got enough materials to start the process and a constant stream of kernels afterwards.

It's find the way it is. It could show a little more information on what each kernels COULD give but the system should not be changed.

Agreeing with Annihilator. NPC nerfs are making research progresively more dificult.

Also why are no prototypes sold on the market ? Anyway semi-specialisation would be good.

F.e. laser weapons and similar class based specialisation. You still have to go through the tiers to reach t4, but now you don't need to waste time and NIC on thins you don't want to.

1. it won't enable to get the complete tech tree any faster, just you can select what goes first.
2. specialising enables to get into higher tier production faster thus more competition on the market

@Cassius: I can understand your point of view as a solo player to have something to shoot for, however it gets more complicated when you try to be competitive as a corp and want to live on beta. To have a real chance to win a serious battle you not only need the manpower, but also the tech. You will loose stuff and must replace it regularly. The only way to do that is to prototype by yourself, buying items is far too expensive and also there is not enough supply.

Right now if you start as a new Corp with 20 active players you would need ages(literally) to get to the point where you can sustain yourself without help from outside. So the only way right now is to join in with veteran players that will help you. It is extremely hard to start something new, much harder than it was a few months ago.

Back when every Bot dropped a kernel the system was ok. After that, it was harder, but still doable. Right now you have to farm much longer to get the kernels you want and higher level indu-kernels are now not farmable at all any more as random spawns do not spawn indu mechs any longer(well they do but only once per day as it seems). So you cant even get proper mining equipment.

Also farming kernels, even with random spawns gets really boring after some time. There should be more ways to gather kernels...maybe let players drop some when they die in PvP?

Well as soon as you got all kernels you need you will stop this stupid grind, wont you? Maybe occasinally kill some bots for the fun of it. So there are no kernels to sell. Instead you will go mining to produce items. You will also only mine as much as you need so you wont sell the items on the market, because at this point you have no use for any nic you might earn.

And Corps do buy kernels...but there are not nearly enough sold, or are sold for insane prices. And how do you get the money to buy kernels these days? Right farm some more NPCs as missions are even worse. And there are also other running costs you have to pay for so the NIC earned while farming kernels is actually needed elsewhere.

Still even if you could buy all the kernels...where do indu kernels come from? And how expensive would they be if someone really did sell them?

I tell you the recent NPC changes really hurt the current system.