AI-Supervised Probabilistic Workflows for Adaptive Automation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current automation systems, such as robotic process automation (RPA), operate in a deterministic fashion with fixed workflows that lack dynamism and adaptability, limiting their effectiveness in dynamic environments.
Innovation Solution
Integrate probabilistic models into deterministic workflows using artificial intelligence (AI) to replace deterministic activities, creating more dynamic and adaptable workflows, and employ a supervisor system to monitor and correct the probabilistic systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If deterministic workflows with set logic are used, then system reliability and predictability are improved, but workflow flexibility and adaptability deteriorate
Solution Approach 1:
The patent applies dynamics by transitioning from static deterministic workflows to dynamic probabilistic workflows. The system uses probabilistic models that can adapt their behavior based on learned patterns and confidence levels, allowing workflows to dynamically adjust their execution path rather than following fixed predetermined rules.
Solution Approach 2:
The patent changes the fundamental parameter of workflow decision-making from deterministic (binary yes/no) to probabilistic (confidence-based). By introducing confidence thresholds and probabilistic outcomes, the system transforms rigid workflow parameters into flexible, adaptive parameters that can respond to varying conditions.
2Adaptability or versatility
If probabilistic models are integrated into workflows, then workflow flexibility and adaptability are improved, but system complexity increases
Solution Approach 1:
The patent introduces an intermediary layer (the probabilistic model and confidence threshold mechanism) between the input data and the workflow execution. This intermediary processes uncertain information and transforms it into actionable decisions, managing complexity by providing a structured interface rather than directly handling raw probabilistic data throughout the system.
Solution Approach 2:
The patent segments the workflow into deterministic segments (where confidence is high) and probabilistic segments (where confidence varies). This allows different parts of the system to operate with appropriate levels of certainty, reducing overall complexity by applying probabilistic methods only where necessary rather than throughout the entire workflow.
3Adaptability or versatility
If AI models replace deterministic activities, then workflow adaptability is improved, but measurement precision and confidence assessment become more difficult
Solution Approach 1:
The patent implements feedback mechanisms where the system continuously monitors confidence levels and adjusts its behavior accordingly. When confidence thresholds are not met, the system can trigger re-evaluation, human intervention, or alternative paths, creating a feedback loop that improves measurement precision through iterative assessment and adjustment.
Data Source
AI summary
Probabilistic models may be used in a deterministic workflow for an automation. Artificial intelligence (AI) introduces a probabilistic framework where the outcome is not deterministic, and therefore, the steps are not deterministic. Deterministic workflows may be mixed with probabilistic workflows, or probabilistic activities may be inserted into deterministic workflows, in order to create more dynamic workflows. A supervisor system may be used to monitor an AI model and raise an alarm, disable an automation, bypass the automation, or roll back to a previous version of the AI model when an error is detected by a data drift detector, a concept drift detector, or both.