Hello,
Using Keystrokes and Mouse clicks you can certainly automate SWING GUI Java-based stand alone applications. You can also use 'Image Recognition' feature that gives Automation Anywhere power to "See the screen". Compare static or dynamic images as task runs and perform actions based on it like a human being would do. It also provides a very powerful feature whereby it allows you to click on the image if it is found within the larger image.
Can you please provide your phone number and best time to reach using following link so a technical support person can call , look into the actual scenario and provide viable solution?
Automation Anywhere - Open A Support Case
Alternatively, you can reach us at 1-888-484-3535 x 3.
Hope that helps.