AI Planning Controller Quiescent Phase Synthesis

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Automated planning in real-time systems, such as IoT systems, faces challenges in adapting to changes in system states and ensuring timely action dispatch, leading to potential delays that can result in unsuccessful task accomplishment.

Innovation Solution

The approach involves identifying a quiescent phase where system states remain stable for plan synthesis and strategically allocating action dispatch to distributed controllers that meet communication latency constraints, ensuring that plan synthesis and action dispatch occur during periods of stability and minimizing delays.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If plan synthesis is performed in real-time systems, then the system can adapt to changes in system states, but delays in plan synthesis may occur leading to unsuccessful task accomplishment

Engineering Contradiction:

Improveadaptability to system state changesVSAvoidplan synthesis delay

Core Design Contradiction:

Adaptability or versatilityVSLoss of time

Solution Approach 1:

The system performs preliminary actions by identifying quiescent phases in advance where system states are predicted to remain stable. Plan synthesis is scheduled to occur during these pre-identified stable periods, allowing the system to prepare and execute planning operations when conditions are favorable, thereby reducing delays while maintaining adaptability to state changes.

Inventive Principle:

Principle #10Preliminary action

2Productivity

If action dispatch is performed by distributed controllers, then the system can meet communication latency constraints, but control signalling latency may cause failed execution

Engineering Contradiction:

Improveaction dispatch efficiencyVSAvoidexecution success rate

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The system employs feedback mechanisms by monitoring communication latencies between distributed controllers and adjusting action dispatch decisions accordingly. Controllers exchange information about their current load and communication status, allowing the system to route control signalling through optimal paths and avoid bottlenecks, thereby maintaining both high productivity and reliable execution.

Inventive Principle:

Principle #23Feedback

3Adaptability or versatility

If plan synthesis occurs during system state changes, then the system can respond dynamically, but the synthesized plan may become immediately stale

Engineering Contradiction:

Improvedynamic response capabilityVSAvoidplan validity

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

The system performs preliminary identification of quiescent phases where system states are predicted to remain stable for sufficient durations. By scheduling plan synthesis operations during these pre-identified stable periods, the system ensures that synthesized plans remain valid and non-stale, while still maintaining the ability to respond dynamically to state changes that occur outside these phases.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS11960926B2Automated plan synthesis and action dispatch

Publication Date: 2024.04.16 TELEFONAKTIEBOLAGET LM ERICSSON (PUBL)

AI summary

An artificial intelligence, AI, planning controller control the timing of when a plan (16) to accomplish a task (14) is synthesized. The AI planning controller in this regard determines a quiescent phase (20) during which values of at least some predicates describing a state of the system (12) will remain stable. The AI planning controller then controls artificial intelligence planning to synthesize the plan (16) during at least some of the quiescent phase (20).