AI-Driven Standby Scheduling for Shared-EV Onboard Computers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing fleet management systems for shared vehicles fail to accurately predict vehicle availability and optimize resource usage, leading to inefficient battery consumption and unnecessary communication, computing, and action request transmission.
Innovation Solution
A method that calculates and adjusts the standby duration, reactivation frequency, and wake-up duration of on-board computer equipment based on vehicle charge state and historical reservation data using artificial intelligence, optimizing battery resources and improving decision-making by the management server.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If computer equipment is kept in standby mode to reduce power consumption, then battery consumption is reduced, but the equipment cannot respond to management server requests in a timely manner
Solution Approach 1:
The patent implements dynamic adjustment of standby duration based on vehicle usage patterns and reservation data. The computer equipment transitions between standby and active states adaptively, with standby duration modified according to predicted vehicle availability. This resolves the contradiction by making the system flexible rather than static, allowing energy savings during low-activity periods while maintaining responsiveness when vehicle usage is anticipated.
Solution Approach 2:
The patent uses historical reservation data and machine learning models to predict vehicle availability in advance. Based on these predictions, the system proactively adjusts the standby duration of computer equipment before the vehicle is actually needed. This preliminary action ensures that equipment is activated just in time for vehicle usage, reducing unnecessary standby periods while avoiding premature activation that would waste energy.
2Reliability
If the management server transmits action requests to all vehicles, then operational status is maintained, but communication and computing resources are wasted on unnecessary transmissions
Solution Approach 1:
The patent applies different communication strategies to different vehicles based on their individual characteristics, location, and predicted usage. Rather than uniform treatment, the system tailors action request transmission to each vehicle's specific context. This resolves the contradiction by allocating communication resources locally and selectively, maintaining operational status where needed while avoiding waste on vehicles unlikely to be used soon.
Solution Approach 2:
The patent dynamically changes communication parameters such as request transmission frequency and timing based on vehicle-specific factors including historical usage patterns, current location, and predicted availability. This parameter adaptation allows the system to optimize communication resource usage while maintaining necessary operational monitoring, transmitting requests more frequently to high-priority vehicles and less frequently to low-priority ones.
3Ease of operation
If action requests are transmitted without considering vehicle waiting times, then management decisions are simplified, but resource allocation efficiency decreases
Solution Approach 1:
The patent implements self-service through automated machine learning models that predict vehicle waiting times and autonomously determine optimal action request timing. The system eliminates the need for complex manual management decisions by automatically analyzing historical data, predicting vehicle availability, and scheduling communications. This resolves the contradiction by transferring decision-making from human operators to an autonomous system, maintaining simplicity for users while achieving high resource allocation efficiency through data-driven predictions.
Solution Approach 2:
The patent incorporates feedback loops where actual vehicle usage data is continuously collected and used to refine waiting time predictions. The system learns from past performance and adjusts its predictions and communication strategies accordingly. This feedback mechanism enables the system to improve resource allocation efficiency over time while keeping management operations simple, as the automated system continuously optimizes based on real-world outcomes.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The invention relates to a process for modifying the standby duration and/or the reactivation frequency and/or the wake duration of on-board devices in electric vehicles belonging to a fleet of shared vehicles located in a defined geographical zone. Said vehicles each comprise an electric engine powered by a rechargeable electric battery and having a standby duration and/or a reactivation frequency and/or a wake duration which are preconfigured. Said process employs a method for calculating the wait time between two reservations of said vehicles, said method comprising the following steps: a) implementing a logical computing process suitable for calculating, for each vehicle of the fleet which is available for reservation, a mean wait time during which said vehicle will not be reserved, said calculation being carried out by executing a computing application based on an artificial intelligence model, b) detecting the battery charge level of each electric vehicle which is available for reservation, c) for each electric vehicle which is available for reservation, weighting the mean wait time calculated in step a) with the battery charge level (or respectively with the tank fill level) of said vehicle so as to obtain a corrected mean wait time for said vehicle. Said process further comprises a step of modifying the standby duration and/or the reactivation frequency and/or the wake duration of the device of the vehicle in question on the basis of the corrected mean wait time assigned to said vehicle.