Intelligent vehicle control via AI-MIMO: Architectural design and simulation-based evaluation
Keywords:
AI-MIMO, Multimodal fusion, Steering angle, Forward lighting, Gear-shift adviceAbstract
This work presents an AI-enabled Multi-Input Multi-Output (AI-MIMO) control architecture aimed at improving safety assistance in automobiles through the utilization of low-cost external sensors that function independently of the vehicle's integrated systems. The suggested architecture fuses three simulated input streams steering angle, forward-lighting intensity, and vehicle speed in real time using MATLAB/Simulink. Fusion is accomplished by a modular pipeline that includes a rule-based preprocessing stage, a lightweight neural-network classifier for estimating the state, and a supervisory controller that manages the system's outputs. The steering angle tells us when to turn, the lighting level tells us when to cut down the glare, and the speed of the car tells us when to change gears. We ran studies that only used simulation, synthetic sensor data, and pre-set driving scenarios to test response time, controller stability, and action correctness. We looked at 30 different scenarios to see how well the quantitative indicators worked. These included the correctness of the turn-signal reminder, the delay in dimming the headlights, and the consistency of the gear-shift advisory. The AI-MIMO architecture produces reliable output coordination, with average action-timing errors below 8% of the scenario duration and correct response rates above 90% for all three tasks. These results show that it is possible to use low-cost sensor inputs and AI-MIMO fusion to provide coordinated safety-assistance actions. However, they are only based on MATLAB simulations and have not yet been tested in real vehicles or hardware-in-the-loop situations. The model will be expanded in subsequent efforts to incorporate physical sensors and embedded implementations.
References
[1] World Health Organization, Global Status Report on Road Safety 2023. Geneva: World Health Organization, 2023.
[2] J. Teknologi et al., “Literature Review : Analisis Pengaruh Teknologi Advanced Driver Assistance System ( ADAS ) dalam Keselamatan Berkendara Analysis of the Influence of Advanced Driver Assistance System ( ADAS ) Technology in Improving Driving Safety Penulis Judul Penelitian,” vol. 15, no. September, pp. 174–183, 2025, doi: 10.34010/jati.v15i2.16718.
[3] M. Azmi, D. Sudarno, W. Purwanto, T. Sugiarto, and D. Fernandez, “Implementasi Jaringan Syaraf Tiruan untuk Mengendalikan Lampu Sein Pada Sepeda Motor,” J. Inov. vokasi dan Teknol., vol. 19, no. 2, 2019, doi: 10.24036/invotek.v19i2.622.
[4] E. Tamakloe et al., “Mobile Robots and Autonomous Vehicle Control : A Comprehensive Review of the Advancements and Challenges,” Makara J. Technol., vol. 29, no. 1, 2025, doi: 10.7454/mst.v29i1.1635.
[5] A. Rizal, “Sistem Kendali Otonom pada Kendaraan Listrik Menggunakan Sensor Fusion dan Kalman Filter,” Jounal Eng. Technol. Sci., vol. 1, no. 2, pp. 68–74, 2025, doi: 10.70716/jets.v1i2.134.
[6] E. N. Budisusila, S. Arttini, D. Prasetyowati, and B. Y. Suprapto, “Neural network training for serial multisensor of autonomous vehicle system,” vol. 12, no. 5, pp. 5415–5426, 2022, doi: 10.11591/ijece.v12i5.pp5415-5426.
[7] F. S. Faizi and A. K. Alsulaifanie, “Steering angle prediction via neural networks,” vol. 31, no. 1, pp. 392–399, 2023, doi: 10.11591/ijeecs.v31.i1.pp392-399.
[8] A. Zaarane, I. Slimani, M. Elhabchi, and I. Atouf, “Enhanced driving assistance : automated day and night vehicle detection system utilizing convolutional neural networks,” vol. 36, no. 3, pp. 1532–1542, 2024, doi: 10.11591/ijeecs.v36.i3.pp1532-1542.
[9] F. Mart, H. Montiel, and F. Mart, “Comparative study of optimization algorithms on convolutional network for autonomous driving,” vol. 12, no. 6, pp. 6363–6372, 2022, doi: 10.11591/ijece.v12i6.pp6363-6372.
[10] A. Mulyanto, R. I. Borman, P. Prasetyawana, and A. Sumarudin, “2D Lidar and Camera Fusion for Object Detection and Object Distance Measurement of ADAS Using Robotic Operating System ( ROS ),” Int. J. Informatics Vis., vol. 4, pp. 231–236, 2020.
[11] R. M. Susanto, T. Nurrochman, S. Munahar, and A. I. Ramadhan, “Automotive Experiences,” J. UnimmaAutomotive Exp., vol. 2, no. 3, pp. 67–72, 2019, doi: doi.org/10.31603/ae.v2i3.2766 Published.
[12] A. Rafi, A. Tahtawi, J. T. Elektro, P. N. Bandung, and K. B. Barat, “Kendali Posisi Motor DC Menggunakan Logika Fuzzy Interval Tipe 2 The Position Controlling of DC Motor Using Interval Type-2 Fuzzy Logic,” TELKAJurnal Telekomun. dan Kontrol, vol. 7, no. 1, pp. 1–10, 2021, doi: 10.15575/telka.v7n1.1-10.
[13] R. S. Saputri, A. Apriliani, R. S. Amar, and L. Y. Astri, “Deteksi Dini Microsleep pada penemudi kendaraan roda empat melalui citra mata menggunakan algoritma YOLOv8,” J. Process., vol. 20, no. 2, pp. 165–179, 2025, doi: 10.33998/processor.2025.20.2.2529.
[14] A. Suhendi and Desmiwati, “RANCANG BANGUN ALAT KEAMANAN PADA KENDARAAN BERMOTOR RODA DUA BERBASIS AI MENGGUNAKAN SENSOR SUARA,” J. Teknol. Inf., vol. 10, no. 2, pp. 168–172, 2024, doi: 10.52643/jti.v10i2.5934.
[15] Y. Away and A. Novandri, Rekayasa Otomasi Cerdas Berbasis Sistem Embedded. Banda Aceh: Syiah Kuala University Press, 2024.
[16] M. M. Abdellatif, M. A. Çifçi, A. A. Ibrahim, H. M. Harb, and A. S. Desuky, “Optimizing internet of things based gas sensors : deep learning and performance optimization strategies,” Int. J. Electr. Comput. Eng., vol. 15, no. 5, pp. 4813–4828, 2025, doi: 10.11591/ijece.v15i5.pp4813-4828.
[17] J. E. Istiyanto, D. Lelono, and O. Natan, “Trainer Kit for Aroma Classification Using Artificial Intelligence,” vol. 13, no. 2, pp. 294–308, 2025, doi: 10.52549/ijeei.v13i2.6146.
[18] B. A. Ampangallo and J. Londongsalu, Keselamatan transportasi. .
[19] M. Zhu, F. Yu, and S. Xiao, “GPS-Integrated Kinematic Positioning and Navigation Method Based on Singer Model,” J. Sens., 2019, doi: doi:10.3390/s19194274.
[20] A. Mofolasayo, “Towards ‘ Vision-Zero ’ in Road Traffic Fatalities : The Need for Reasonable Degrees of Automation to Complement Human,” J. Syst., 2024, doi: 10.3390/systems12020040.
[21] U. Brawijaya, “PENERAPAN DEEP CONVOLUTIONAL GENERATIVE ADVERSARIAL NETWORK UNTUK MENCIPTAKAN DATA SINTESIS PERILAKU PENGEMUDI DI DALAM KENDARAAN IMPLEMENTATION OF DEEP CONVOLUTIONAL GENERATIVE ADVERSARIAL NETWORK TO GENERATE SYNTHETIC DATA OF DRIVER ACTIVITY INSIDE,” vol. 10, no. 5, pp. 963–972, 2023, doi: 10.25126/jtiik.2023106978.
[22] D. D. Thu, “MODELING AND SIMULATION OF GEAR SHIFT CONTROL FOR THE U340 AUTOMATIC TRANSMISSION USING MATLAB–SIMULINK AND STATEFLOW,” Int. J. Appl. Math., vol. 38, no. 12, pp. 120–132, 2025, doi: https://doi.org/10.12732/ijam.v38i12s.1362.
[23] O. Galchonkov, M. Babych, A. Zasidko, and S. Poberezhnyi, “USING A NEURAL NETWORK IN THE SECOND STAGE OF THE ENSEMBLE CLASSIFIER TO IMPROVE THE QUALITY OF CLASSIFICATION OF,” pp. 15–21, 2022, doi: 10.15587/1729-4061.2022.258187.
[24] V. Slyusar et al., “CONSTRUCTION OF AN ADVANCED METHOD FOR RECOGNIZING MONITORED OBJECTS BY A CONVOLUTIONAL NEURAL NETWORK USING A DISCRETE WAVELET,” pp. 65–77, 2021, doi: 10.15587/1729-4061.2021.238601.
Downloads
Published
Conference Proceedings Volume
Section
License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.