Cloud-Hosted Automated Call Assistant for Low-Battery Reliability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Local implementation of automated assistants for handling telephone calls consumes computational and battery resources at client devices, and may fail when devices have low battery, without user awareness, and fails to manage unwanted calls effectively.
Innovation Solution
A cloud-based automated assistant handles incoming calls remotely, determining actions based on user and device states, and managing calls without local resource consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If automated assistants are implemented locally at client devices, then they can handle telephone calls independently, but they consume computational resources and battery resources at the client device
Solution Approach 1:
The automated assistant functionality is extracted from the client device and relocated to a remote server. The client device only needs to route incoming calls to the server, while the computationally intensive speech recognition, natural language understanding, and text-to-speech processing are performed remotely, eliminating local battery consumption while maintaining call handling capability
Solution Approach 2:
A remote server acts as an intermediary between the client device and the automated assistant functionality. The server receives call routing requests from the client, processes the automated assistant operations, and returns results to the client, thereby enabling call handling without direct local resource consumption
2Extent of automation
If automated assistants run as background processes, then they can answer incoming calls automatically, but users may not be aware of the resource consumption
Solution Approach 1:
The system provides feedback to users about automated assistant operations through notification messages. When the automated assistant answers a call or performs actions, the system sends notifications to the user's device, making users aware of the automated operations and associated resource usage without requiring continuous monitoring
3Productivity
If automated assistants are implemented locally, then they can process calls independently, but they fail when the client device has less than a threshold state of charge
Solution Approach 1:
By extracting the automated assistant from the client device and hosting it on a remote server with adequate power supply, the system eliminates the dependency on client device battery charge levels. The server can maintain continuous operation regardless of the client's state of charge, ensuring reliable call processing
Solution Approach 2:
The problem is solved by moving the automated assistant operation to a different dimensional context - from the constrained client device environment to the unconstrained server environment. This dimensional shift allows the system to operate independently of the client device's power state while maintaining full call processing capability
4Adaptability or versatility
If automated assistants handle all incoming calls, then they can manage conversations, but they cannot effectively filter unwanted calls like telemarketers and spammers
Solution Approach 1:
The system performs preliminary analysis of incoming calls before full automated assistant engagement. Call screening mechanisms analyze caller information, patterns, and content in advance to identify and filter unwanted calls from telemarketers and spammers, preventing them from reaching the automated assistant and wasting resources
Solution Approach 2:
Different quality levels of processing are applied to different calls. Wanted calls receive full automated assistant conversation management, while unwanted calls are filtered out with minimal processing. This local quality differentiation optimizes resource allocation and effectively manages harmful factors
Data Source
AI summary
Implementations are directed to a cloud-based automated assistant that can handle incoming telephone calls on behalf of a user. In some implementations, processor(s) of a remote server can receive an indication of an incoming telephone call that is directed to a client device and via a cloud-based automated assistant executing at least in part at the remote server (e.g., that is remote from the client device). The processor(s) can determine a next action to be implemented by the cloud-based automated assistant and cause the next action to be implemented by the cloud-based automated assistant. In some implementations, the indication of the incoming telephone call can be directed to the cloud-based automated assistant (e.g., without the client device routing the incoming telephone call to the cloud-based automated assistant). In other implementations, the indication of the incoming telephone call is routed to the cloud-based automated assistant via the client device.