I have a large array of use cases where throughout my shift as a project manager I want to be able to use ai to assist with these three scenarios:
1. Taking a quick hand drawn sketch and turning it into a professional looking diagram to explain concepts, work status and required tasks.
- Using a photo or photos taken of different areas on the processing plant of the mine and overlaying a hand drawn sketch or a description to illustrate work that has been done, work that will need to be completed, safety issues or ideas for improvement.
-
Using just a promot to create technical drawings or diagrams to explain concepts and ideas.
I am having difficulty having the image generation tools I've used online so far as well as nano banana and chatgpt to maintain the sketched ideas layouts and specific information.
I have some examples.
I'm trying to depict that the setup for changing out a conveyor belt has been secured in a way that is dangerous and poses a risk to workers. I have a sketch mock-up I quickly did on my phone that is colour coded and am utilising this prompt:
"This image is a side on profile sketch depicting a conveyor system with a GTU. In this sketch is a depiction of the current state of a conveyor belt change out job that is underway. The blue and red lines represent the conveyor belt. The green crosses are depicting belt clamps that are correctly setup and respectfully depict a compliant belt clamp on the carry side of the conveyor belt immediately next to the tail pulley, and a compliant belt clamp immediately prior to the head pulley on the carry side, these follow safe practises that ensure the conveyor belt can not move between them. This is why the conveyor belt that is between them is depicted as white for safe and secured with no chance of movement due to correct clamping. The orange/yellow crosses depict non compliant clamps in place that are securing the two ends of the conveyor belt as it is folded back onto itself to expose the space that the vulcaniser press that will be used to splice the two ends of the belt together will be setup. This folded back part of the belt is depicted in red as it is secured with non compliant belt clamps so it can not be assumed that the stored energy contained in the belt will not be released resulting in the belt folding back over the exposed section and resting naturally with the ends overlaid in the middle. The red arrows depict this potential movement. The danger sign symbol depicts the danger zone for any personal that are within the line of fire of this potential movement.
Please turn this into a professional looking diagram that clearly depicts this. Include all of the above mentioned features as well as pulley locations (head, tail, gtu entry, gtu & gtu exit), the stored energy locations, include labels."
The results so far have not been usable and are moving critical parts of the layout that eliminate the purpose of the generated image by not maintaining the core principles that are being depicted.
(Moving the GTU location, creating non sense lines of the conveyor belt, depicting energy sources and potential movement in nonsense directions etc.)
The other example is I am trying to get the tool to sketch up a required scaffolding onto a photo I have taken. I have had no luck with this either.
Would greatly appreciate any suggestions for tools, process's and prompting structure to help with these tasks.