Vehicle Range Estimation Using ML Kernel Predictors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional systems and methods fail to accurately predict the energy consumption of performing AI algorithms in autonomous vehicles, leading to inaccurate remaining range estimates for vehicle drivers.
Innovation Solution
The system employs a plurality of predictors and a machine-learning model comprising kernels to estimate the energy consumption of running machine-learning models on vehicle hardware devices, and subsequently calculates the remaining vehicle range based on this energy consumption and vehicle-specific information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional systems and methods are used to estimate remaining range, then the system is simple, but the measurement precision of energy consumption prediction is poor
Solution Approach 1:
The system segments the energy consumption prediction task by dividing it into multiple predictors, each responsible for estimating energy consumption of specific machine-learning models or hardware devices. This segmentation allows for more precise predictions while maintaining manageable system complexity through modular architecture.
Solution Approach 2:
The patent introduces predictors as intermediary components that bridge the gap between machine-learning model execution and energy consumption estimation. These predictors act as mediators that translate model characteristics into accurate energy consumption predictions, thereby improving measurement precision without directly increasing the complexity of the core ML infrastructure.
2Productivity
If AI algorithms and on-vehicle sensors are performed, then the functionality and task performance are improved, but the energy consumption increases
Solution Approach 1:
The system performs preliminary estimation of energy consumption using multiple predictors before executing machine-learning models. This preliminary action allows the vehicle to predict energy requirements in advance, enabling better energy management and range estimation without compromising the actual task performance of the AI algorithms.
Solution Approach 2:
The patent implements a feedback mechanism where energy consumption predictions from multiple predictors are continuously updated and used to refine remaining range estimates. This feedback loop allows the system to adapt to actual energy consumption patterns while maintaining accurate productivity assessments of AI task performance.
3Reliability
If accurate remaining range estimation is provided, then the reliability of driver decision-making is improved, but the measurement precision requirements increase
Solution Approach 1:
The system merges predictions from multiple predictors to produce a comprehensive energy consumption estimate. By combining multiple prediction sources, the system achieves high measurement precision required for reliable driver decision-making while distributing the precision requirement across multiple components rather than placing the entire burden on a single measurement system.
Data Source
AI summary
A system includes a controller configured to determine a machine-learning model comprising kernels among a plurality of machine-learning models based on a task to be performed by a vehicle, select one of a plurality of predictors based on a hardware device of the vehicle, each of the plurality of predictors predicting energy consumption of the kernels in corresponding hardware device, estimate, using the selected predictor, energy consumption of running the machine-learning model for performing the task on the hardware device of the vehicle, and estimate remaining range of the vehicle based on the estimated energy consumption, and information of the vehicle, and a route of the vehicle.