I have finished a new method (implemented as an algorithm/software) which exploring data inputs obtains a sequence of steps, appliable for different industrial processes, that reduces drastically material and human resources as well as energy and time.

For example. This method gives the optimal commands for managing an arm-robot to arrange 7 tones of different size stones in 10 pallets of 1 cubic metre in 60 minutes instead of the current process that needs 3 arm-robots (and others) for performing the same task in 30 pallets and in more than 300 minutes.

Note that for different processes and inputs (a different set of stones) it will produce a different sequence of commands, so it is generating new industrial processes for each new task.

I know that software that implements industrial processes are patentable in US and EU but as I´ve exposed it is not an industrial process but an algorithm to generate some industrial processes.

Is this case of software patentable as a unique-global software for industrial processes or each case must be patented apart?

2 Answers 2


A U.S. answer. Some things in the field you describe are patentable in the U.S. and some are not. Unfortunately, above novelty and non-obviousness, the current huge hurdle is abstractness. The law on this in the U.S. has changed in the last few years in the direction of making it much easier to shoot down something as abstract. The broader range of processes you intend to cover the more likely that will trip you up.

It takes a very skilled patent attorney or agent to draft a specification and claims to avoid the abstractness black hole. All USPTO registered practitioners (and those in other locations) need to have at least an undergraduate scientific or engineering background, unlike generic attorneys. I have met several PhD patent attorneys. You can find a practitioner who is expert in your field who has successfully prosecuted other's patents.

In U.S. law it used to be that a programed computer was seen as a machine (patentable) and as a different machine than an unprogramed or differently programed machine. Now, unfortunately it is often seen as a computer "just doing what computers do". Yes this is ridiculous. A good practitioner can couch your invention in terms that make it tangible. Maybe they do show a hardwired version as an example embodiment but have some claims that are more general. The system is illogical but there is often a way through it.

  • Thank you George ... It is true that many programs are not inventions, only different expressions, but a program as an algorithm expression can decipher the DNA or predict the weather of tomorrow ... these algorithms can never be performed by a mechanism , only for computers or electronic devices, at least for now. I understand that my invention is of this type and therefore its patent is justifiable. Do you?
    – Izar Urdin
    Commented Feb 11, 2020 at 5:59
  • It depends on the details of the invention and the jurisdiction.
    – George White
    Commented Feb 11, 2020 at 16:32

My answer refers to European patent practice. Most probably you will not be able to get a patent granted for the method/algorithm if you do not refer it to a particular industrial application. Without claiming the different steps of the industrial procedure, the steps of your method/algorithm will be considered a mathematical method or a computer program as such (depending on how it the claims are drafted), both of which are not regarded as inventions.

An algorithm is in fact patentable when it provides a technical effect, which can also be seen as solving a technical problem. It is true that your algorithm enables you to solve a technical problem whereby fewer resources/time/etc. are necessary, but as long as you do not set out the actual industrial process in the claim the algorithm is just an algorithm, not a solution to a technical problem. The second alternative you mention is, in my opinion, the one that applies in your situation: you have to patent the different cases. Before filing for a patent, it would be wise to study the different scenarios that your technology may address and have them all included in the same patent application or file different patent applications.

  • It seems really absurd to think that an algorithm is not a technological solution when it is precisely the only way to obtain the expected result. This same algorithm could be implemented with transistors and then it would be patentable, right?
    – Izar Urdin
    Commented Feb 9, 2020 at 18:31
  • 1
    @IzarUrdin I know what you mean, but I am afraid it would not make it patentable. You could even take it one step further and claim a computer or a computer program with instructions to carry out the algorithm, because those are patentable whenever they solve a technical problem. Unless I have not got the details from your question properly, your algorithm provides you with a sequence of operating something in an efficient way, but you can use that algorithm in several devices/industries under certain conditions. You have to claim those devices/conditions for the algorithm to be patentable. Commented Feb 9, 2020 at 22:11
  • 1
    Great answer for EPO. Just a note that the need for a patentable invention to be a technical solution to a technical problem is true in EPO and most of the world but not in the U.S. For example many forms of games and "novelties" have been patentable in the U.S.
    – George White
    Commented Feb 9, 2020 at 22:48

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .