Configure AI-Assisted Self-Healing Robots
Engage AI at the point of failure and get possible resolutions as AI recommendations, apply the recommendations to the failed action and recover successfully in runtime and push these recommendations to design time to be adopted in the future robot versions.
Robots are fragile and tend to fail due to minor changes in the UI. UI elements might change frequently: IDs, class names, structure, or alignment may vary across deployments or sessions. Such changes often break recorded targets.
AI-assisted self-healing enables robots to automatically recover from certain UI-related failures during runtime. An AI-assisted self-healing robot improves the reliability when minor or incremental UI changes are implemented. When a robot cannot locate or interact with an element, the system analyzes the issue and attempts to resolve it using AI-generated recommendations.
At runtime, AI-assisted self-healing allows robots to:
- Automatically recover from UI-related failures
- Adapt to minor application changes
- Continue execution with minimal interruption
- Generate AI recommendations which can be verified and incorporated at design time to create a new version of the robot.
AI-assisted self-healing enables robots help you in the complete lifecycle from error discovery, recovery and making it available at design time for creating a new version with the AI recommendations applied. This feature helps reduce manual maintenance and improves the reliability of robot workflows.
For example, if there are modifications in XPath or CSS selectors, the AI-assisted seal-healing feature will recommend the possible alternatives by analyzing the error, in context of the automation, using recommendations from Oracle Cloud Infrastructure Generative AI (OCI Generative AI).
AI-assisted seal-healing then takes corrective action by applying the recommendations to successfully recover the robot from the error and proceed with the rest of the automation without requiring changes to the original configuration.