search

OpenAdaptAI / OpenAdapt

AI-First Process Automation with Large ([Language (LLMs) / Action (LAMs) / Multimodal (LMMs)] / Visual Language (VLMs)) Models
https://www.OpenAdapt.AI
MIT License
764 stars 104 forks source link

Usecase: aerospace and aviation #154

Open abrichr opened 1 year ago

abrichr commented 1 year ago

How can we enable the aerospace and aviation industry to automate repetitive tasks?

This involves:

  1. Identifying specific use cases relevant to the aerospace and aviation industry (i.e. job positions, associated tasks, and relevant software applications)
  2. Creating recordings involving one or more relevant applications, and implementing tests for corresponding completions:
KrishPatel13 commented 1 year ago

To enable the aerospace and aviation industry to automate repetitive tasks, here are some examples that involves repitative tasks:

  1. Use Case: Aircraft Design and Engineering Job Positions: Aerospace Engineers, Designers Tasks: Creating 2D and 3D models, drafting blueprints, performing structural analysis Relevant Software Applications: Autodesk AutoCAD, Siemens NX, CATIA

  2. Use Case: Maintenance and Inspection Job Positions: Aircraft Maintenance Technicians, Inspectors Tasks: Checking and recording aircraft components, conducting inspections, identifying defects Relevant Software Applications: Honeywell Forge, Skywise, Ramco Aviation

  3. Use Case: Flight Operations and Air Traffic Management Job Positions: Air Traffic Controllers, Dispatchers Tasks: Flight planning, route optimization, monitoring aircraft performance Relevant Software Applications: FlightAware, Aireon, Skyplan

Next Steps:

  1. A common repititive task that can be automated is the generation of 2D orthographic views from a 3D model. Traditionally, engineers and designers would have to manually create multiple 2D views (such as top, front, side views) of a 3D model to be used in drafting blueprints. However, with the automation capabilities of software applications, this process can be streamlined.
KrishPatel13 commented 1 year ago

To implement the Step 1, we can break down the process into 2 sub-steps:

  1. Create screen recordings or demonstrations showcasing the relevant software applications and how they automate or assist in performing the tasks.
  2. Implement tests for corresponding completions by providing sample input data or scenarios and asking the AI to generate the desired automated outputs or actions.
KrishPatel13 commented 1 year ago

I have downloaded AutoCAD software and I am exploring for 3D -> 2D Orthographic views modelling as this is a common task as illustrated in https://github.com/MLDSAI/OpenAdapt/issues/154#issuecomment-1551899852

Currently I am not able to record https://github.com/MLDSAI/OpenAdapt/issues/148, I am trying my best to figure out the work-around.

KrishPatel13 commented 1 year ago

We would need a 3D model that is generally being used in the industry as a starting point for the 2D modelling

On researching online, there are almost no free 3D aircraft-related models (DVG) available. Those which look good at a glance are paid.

Moreover, we would need a specific format of the 3D model: DVG. As AutoCAD support importing pre-built 3D model of DVG format.

KrishPatel13 commented 1 year ago

I have downloaded AutoCAD software and I am exploring for 3D -> 2D Orthographic views modelling as this is a common task as illustrated in #154 (comment)

Currently I am not able to record #148, I am trying my best to figure out the work-around.

Update: I am able to record now! :)

Next Steps: