Hi ​@Melqui 4737 ,

Here are some potential causes and solutions:

1. Object properties changed: The object properties of the textbox element might be different on the unattended machine compared to where you developed the bot.
   • Solution: Recapture the object on the unattended machine or use more reliable selectors.
2. Timing issues: The page might not be fully loaded when the bot tries to interact with the textbox.
   • Solution: Add a wait action before the step or use a wait for condition.
3. Browser version differences: Different Chrome versions might render elements differently.
   • Solution: Make sure both environments use the same Chrome version.
4. Resolution/scaling differences: Screen resolution or scaling can affect element recognition.
   • Solution: Set the same display settings on both machines.
5. Selector strategy: The selector might be too specific or using unreliable attributes.
   • Solution: Try using different selector strategies like XPath, CSS, or object cloning.
6. iFrame issues: If the textbox is inside an iFrame, the bot might not be accessing it correctly.
   • Solution: Add a step to explicitly switch to the correct iFrame.
7. Hidden elements: Sometimes elements are initially hidden and become visible after JavaScript execution.
   • Solution: Add a condition to check if the element is visible before interacting.

To troubleshoot this specific issue, I recommend:

1. Run the bot in debug mode to see exactly where it fails
2. Consider using more robust selectors (try CSS or XPath)
3. Add sufficient waits between steps
4. Use the "Recorder" feature to recapture the object on the unattended machine

If there are any differences between the bot development machine and the bot runner machine, you can occasionally see these problems. Many companies use VMs for bot building to keep them identical to the VMs for bot runners, just to avoid these kinds of issues.

Also watch for sites that use dynamic properties for objects, such as IDs.