Vision Builder AI - The easy choice for MOST applications.
Pros
- The interface is extremely easy to use. This makes development very fast, which is key to us. It also makes adjustments on the deployed system easy.
- The deployment is easy because the code you write and test with can be converted to a deployed solution (with UI) in the click of button.
- It is agnostic to the hardware. You can use any machine vision camera on any PC.
- There's a state machine interface for architecting complex applications without having to do the process of converting from a state chart/flow diagram to code.
Cons
- Vision Builder AI is for serial processes, which means there really isn't any parallel processing happening. This works great for a production line where things are coming down in a row but isn't suited for applications where the processing needs to branch and happen in parallel with another process.
- There is code generation built-in if you want to go from Vision Builder AI to LabVIEW but it's got so many wrappers around it that many times I find it faster and easier to just write it all from scratch in LabVIEW (if you want to go that route) rather than doing the code generation.
- If I create a pattern matching template in LabVIEW or the stand-alone template editor I can't seem to use that in Vision Builder AI because it doesn't give me the option to load one from file. I have to create the template in Vision Builder AI.
Return on Investment
- Since machine vision development is such an iterative process, the ease of use of the software cuts down development time significantly.
- Many times we don't have to create a custom solution for the machine vision application. We can just use and deploy Vision Builder AI out of the box.
- The simplified software development provided by Vision Builder AI means we can spend more of our time on the mechanical side.
Alternatives Considered
National Instruments LabVIEW
Other Software Used
TeamViewer, Notepad++