User Power
Value/Post Ratio
128%
- Oct 25, 2018
- 43
- 55
There is one business idea I can execute with the help of my knowledge on machine learning and coding.
I am working as a mechanical engineer and have noticed that many of our CAD designers are losing a lot of time with repetitive tasks that can be written down in an algorithm.
Hence, there is a high potential to automate repetitive, low-creativity tasks that just take too much time and keep motivation to minimum.
Unfortunately, this is a very bad idea with respect to control, because such an automation code is solely customized in the commercial CAD software a company uses and the specific tasks the company repeats often for their typical projects.
Developing and selling such a code can only have two alternatives:
1) specific code for each company using the SDK of the commercial CAD software (cannot be automated and replicated)
2) selling the machine learning code would mean that I compete against the commercial CAD software companies and their research (not realistic)
Hence, the next best thing would be to sell it to my employer, since I know the specifics we need to customize for our daily business.
Implementing it would mean ca 10 working days from me and 20 working days for a programmer from our department.
It would also mean that I would apply knowledge that is beyond my job description (mechanical engineers are not required to know how to develop and code machine learning).
The profit out of it would be freeing up 1 hour of work for every CAD designer for at least 50% of the work of a week in the department for the rest of our lives. Having 60 CAD designers, that can exponentially grow beyond the break even point within 8 weeks.
How could I best go about to make a profit if I sell it only to my employer?
Asking money for it could create a negative impression, if not presented in a smart manner.
I would do all of this for an extra 1,000 € per month, but not for less.
Now that I’m writing it down, it seems like it’s not worth the effort.
I am working as a mechanical engineer and have noticed that many of our CAD designers are losing a lot of time with repetitive tasks that can be written down in an algorithm.
Hence, there is a high potential to automate repetitive, low-creativity tasks that just take too much time and keep motivation to minimum.
Unfortunately, this is a very bad idea with respect to control, because such an automation code is solely customized in the commercial CAD software a company uses and the specific tasks the company repeats often for their typical projects.
Developing and selling such a code can only have two alternatives:
1) specific code for each company using the SDK of the commercial CAD software (cannot be automated and replicated)
2) selling the machine learning code would mean that I compete against the commercial CAD software companies and their research (not realistic)
Hence, the next best thing would be to sell it to my employer, since I know the specifics we need to customize for our daily business.
Implementing it would mean ca 10 working days from me and 20 working days for a programmer from our department.
It would also mean that I would apply knowledge that is beyond my job description (mechanical engineers are not required to know how to develop and code machine learning).
The profit out of it would be freeing up 1 hour of work for every CAD designer for at least 50% of the work of a week in the department for the rest of our lives. Having 60 CAD designers, that can exponentially grow beyond the break even point within 8 weeks.
How could I best go about to make a profit if I sell it only to my employer?
Asking money for it could create a negative impression, if not presented in a smart manner.
I would do all of this for an extra 1,000 € per month, but not for less.
Now that I’m writing it down, it seems like it’s not worth the effort.
Dislike ads? Remove them and support the forum:
Subscribe to Fastlane Insiders.