Travel Directions Desktop Automation Robot on macOS

Cross-platform desktop automation using image template matching and keyboard shortcuts.


In image template-based desktop automation, you provide the robot with screenshots of the parts of the interface that it needs to interact with, like a button or input field. The images are saved together with your automation code. The robot will compare the image to what is currently displayed on the screen and find its target.

Cross-platform desktop automation library

Robocorp provides cross-platform desktop automation support with the RPA.Desktop library. It works on Windows, Linux, and macOS.

Travel directions robot

This example robot demonstrates the use of image templates and keyboard shortcuts to find travel directions between two random locations on Earth.

The robot:

  • Interacts with a web browser to select two random locations on Earth (from
  • Tries to find the directions using the Maps desktop app on macOS (Monterey), using image templates and keyboard shortcuts.
  • Falls back on the web version of Google Maps if Maps fails to find directions.

Note: This robot requires macOS Monterey. (Tested on version 12.4.) The layout and the behavior of the Maps app vary between macOS releases. macOS will ask for permissions the first time you run the robot. Go to System Preferences -> Security & Privacy and check Robocorp Lab, Code, VS Code or Terminal (depending on where you run the robot from) in the Accessibility and Screen Recording sections.

Another important topic:

System settings can impact the recognition of the images: How the interface elements look on a screen depends on system settings like color schemes, transparency, and system fonts. Images taken on a system might end up looking different than the target system, and the robot might not recognize them, stopping the process.

In this case, macOS should use the "Dark" appearance under System Preferences -> General. See our Desktop automation page for more information.

The settings

The robot uses three libraries to automate the task. Finally, it will close all the browsers it happened to open.

The task: Find travel directions between two random locations


Keyword: Get random locations

The robot uses a web browser to scrape and return two random locations from a suitable website.

Keyword: Open the Maps app

The robot opens the Maps app using the Run Process keyword from the Process library. It executes the open -a Maps command. You can run the same command in your terminal to see what happens!

The robot knows when the Maps app is open by waiting for the Maps.MapMode image template to return a match.

Keyword: Maximize the window

The robot maximizes the Maps app window using a keyboard shortcut unless the app is already maximized. The Run Keyword If is used for conditional execution.

The robot knows the Maps app is maximized when the Desktop.WindowControls image template does not return a match (when the close/minimize/maximize icons are not anywhere on the screen).

Keyword: Open and reset the directions view

The robot sets the directions view in the Maps app to a known starting state (empty from and to locations).

  • Conditional execution is used to handle the possible states for the view (it might or might not be open already).
  • Image templates are used to wait for specific app states so that the robot knows when something has been completed.
  • Keyboard shortcuts are used to toggle the directions view.

Keyword: View directions using Google Maps

The robot waits until Google Maps has loaded the directions and takes a full web page screenshot.

Keyword: Enter location

The robot needs to input the from and to locations. This keyword provides a generic way to target those elements on the UI.

Keyword: View directions

The robot tries to find the directions using the Maps app. If that fails, the robot gets the directions from Google Maps.


  • Image template matching is a cross-platform way to find and target UI elements.
  • Keyboard shortcuts are the preferred way to interact with desktop applications (the shortcuts are usually more stable and predictable than the UI).
  • Conditional logic can be used to select different actions based on the state of the application.

Technical information

Last updated

July 7, 2022


Apache License 2.0