Comprehensive Evaluation of the AMAIA Algorithm Efficiency in Predictive Control for Autonomous Driving and Its Industrial Application
Year: 2025
Type: article
Abstract: Context and relevance: The rapid advancement of embedded systems, Internet of Things (IoT), and edge computing technologies has led to an increasing demand for microprocessor architectures capable of operating efficiently under multiple dynamic and energy-constrained environments, Traditional control and optimization techniques often fall short of balancing the simultaneous requirements of energy efficiency, computational performance, and real-time adaptability, particularly in systems facing heterogeneous workloads and changing operational contexts, Objective:T he study aims to present AMAIA (Adaptive Machine Intelligence for Architecture), a novel algorithm grounded in machine learning principles, designed to optimize embedded microprocessor systems by dynamically controlling and predicting system behavior to improve energy efficiency and computational stability, Hypothesis: It is hypothesized that the AMAIA algorithm, through its adaptive and predictive capabilities, can outperform traditional control methods in embedded environments by achieving higher accuracy, reduced error rates, and enhanced energy performance, particularly in scenarios with uncertain and fluctuating constraints, Methodology: The AMAIA framework incorporates Model Predictive Control (MPC) to enable real-time adaptation of control signals based on system states, predictive modeling, and predefined optimization goals,Dynamic voltage and frequency scaling (DVFS) is integrated to reduce energy use, while workload forecasting is performed using ARIMA time-series models, System robustness is maintained using Lyapunov-based stability analysis, Experimental evaluations were carried out by benchmarking AMAIA against traditional PID controllers in real-world driving scenarios (urban and highway conditions),Additionally, the algorithm was deployed in various industrial contexts including autonomous vehicles, industrial process automation, and real-time medical wearable systems., Results: The application of AMAIA yielded notable improvements in control performance, demonstrating a 36% reduction in Mean Absolute Error (MAE), a 61% decrease in Root Mean Square Error (RMSE), and a 31% gain in energy efficiency compared to baseline controllers, statistical validation using paired t-tests (p < 0.001) confirmed significant enhancements in system tracking accuracy, settling time (2.12 s vs. 3.45 s), and control signal smoothness,In industrial deployments, AMAIA achieved an 18.6% reduction in energy consumption and maintained real-time response with processing latency under 30 ms in medical bio signal analysis, Conclusions: The study establishes the AMAIA algorithm as an effective, scalable, and energy-aware solution for embedded microprocessor systems operating in dynamic environments, by integrating predictive control, adaptive system response, and energy-conscious strategies, AMAIA contributes meaningfully to the field of adaptive control and provides a robust framework for enhancing performance in mission-critical domains such as autonomous mobility, industrial process control, and medical technology.
Source: Modelling and Data Analysis
Authors: A, Ali, E.V. Lyapuntsova
Language: English
Subfield: Control and Systems Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7082641316