Neuro-Symbolic Small Language Models (NS-SLM) for Zonal Fault Diagnosis
DOI:
https://doi.org/10.63282/3050-9416.IJAIBDCMS-V7I2P124Keywords:
Small Language Models (SLMs), Neuro-Symbolic AI, Zonal Architecture, Software-Defined Vehicles (SDV), Mixed-Criticality Virtualization, AUTOSAR XML, Knowledge Graphs, ISO 26262, Edge Computing, Fault DiagnosisAbstract
The move towards Software-Defined Vehicles (SDVs) and zonal electrical/electronic (E/E) architectures in the automotive industry requires localized, intelligent edge computing to handle increasing diagnostic complexities. This study introduces a new framework that uses Neuro-Symbolic Small Language Models (NS-SLM) optimized for vehicle zonal gateways. By combining a quantized SLM focused on neural pattern recognition with a formal Knowledge Graph built from AUTOSAR XML topological data offering symbolic, deterministic logicthis hybrid approach delivers highly accurate, explainable fault diagnosis with nearly zero hallucinations. Implemented through mixed-criticality virtualization on hardware such as the ARM Cortex-R52, this design balances the adaptability of generative AI with the stringent ISO 26262 safety standards.
References
1. Toto, G. A., & Limone, P. (17 April 2026). Psychological Dimensions Involved in Image Communication: A Multidisciplinary Research Proposal for Analyzing Cognitive and Perceptual Processes in Visual Education. Proceedings, 139(1), 7. https://doi.org/10.3390/proceedings2026139007
2. Khiabani, Y. S., Atif, F., Hsu, C., Stahlmann, S., Michels, T., Kramer, S., Heidrich, B., Sarfraz, M. S., Merten, J., & Tafazzoli, F. (2025). Optimizing Small Language Models for In-Vehicle Function-Calling. ArXiv. https://arxiv.org/abs/2501.02342
3. Hussain, Md & Rahman, Md & Devarajan, Ramasamy. (2024). Artificial Intelligence-Driven Vehicle Fault Diagnosis to Revolutionize Automotive Maintenance: A Review. Computer Modeling in Engineering & Sciences. 141. 951-996. 10.32604/cmes.2024.056022.
4. Khiabani, Y. S., Atif, F., Hsu, C., Stahlmann, S., Michels, T., Kramer, S., Heidrich, B., Sarfraz, M. S., Merten, J., & Tafazzoli, F. (2025). Optimizing Small Language Models for In-Vehicle Function-Calling. ArXiv. https://arxiv.org/abs/2501.02342
5. Huang, Y., Zhan, R., Wong, D. F., Chao, L. S., & Tao, A. (2025). Intrinsic Model Weaknesses: How Priming Attacks Unveil Vulnerabilities in Large Language Models. ArXiv. https://arxiv.org/abs/2502.16491
6. Huang, Y., Zhan, R., Wong, D. F., Chao, L. S., & Tao, A. (2025). Intrinsic Model Weaknesses: How Priming Attacks Unveil Vulnerabilities in Large Language Models. ArXiv. https://arxiv.org/abs/2502.16491
7. Hussain, Md & Rahman, Md & Devarajan, Ramasamy. (2024). Artificial Intelligence-Driven Vehicle Fault Diagnosis to Revolutionize Automotive Maintenance: A Review. Computer Modeling in Engineering & Sciences. 141. 951-996. 10.32604/cmes.2024.056022.
8. Derse, C., Chakraborty, S., & Hegazy, O. (Mar 2026). Recent Techniques Used for Anomaly Detection in the Automotive Sector: A Comprehensive Survey. Applied Sciences, 16(5), 2584. https://doi.org/10.3390/app16052584
9. Q. Zhang, Z. Liu and S. Pan, "The Rise of Small Language Models" in IEEE Intelligent Systems, vol. 40, no. 01, pp. 30-37, Jan.-Feb. 2025, doi: 10.1109/MIS.2024.3517792.
10. Alansari, A., & Luqman, H. (2025). Large Language Models Hallucination: A Comprehensive Survey. ArXiv. https://arxiv.org/abs/2510.06265
11. Semerikov, S. O., Vakaliuk, T. A., Kanevska, O. B., Ostroushko, O. A., & Kolhatin, A. O. (2025). Edge intelligence unleashed: A survey on deploying large language models in resource-constrained environments. Journal of Edge Computing, 4(2), 179–233. https://doi.org/10.55056/jec.1000
12. Jang, S., & Morabito, R. (2025). Edge-First Language Model Inference: Models, Metrics, and Tradeoffs. ArXiv. https://arxiv.org/abs/2505.16508
13. Kalimuthu, Naresh. (Jan 2026). Small Language Models and Neuro-Symbolic AI in Zonal Architectures: The Rise of Small Language Models (SLMs) in Constrained Environments. International Journal of Emerging Trends in Computer Science and Information Technology. 7. 285-289. 10.63282/3050-9246.IJETCSIT-V7I1P141.
14. R. Fitas, "Neuro-Symbolic AI for Advanced Signal and Image Processing: A Review of Recent Trends and Future Directions," in IEEE Access, vol. 13, pp. 143360-143376, 2025, doi: 10.1109/ACCESS.2025.3598909.
15. Arachchige, Prashani & Iancu, Bogdan & Lilius, Johan. (2025). A Roadmap Toward Neurosymbolic Approaches in AI Design. IEEE Access. PP. 1-1. 10.1109/ACCESS.2025.3617771.
16. A. Santos, J. Martins, J. Sousa, M. Rodríguez and S. Pinto, "Let’s Get Physical: Rethinking the Static Partitioning Hypervisor Architecture for an MMU-Less Memory Model," in IEEE Access, vol. 13, pp. 206046-206065, 2025, doi: 10.1109/ACCESS.2025.3636061.
17. Lu, Q., Li, R., Sagheb, E., Wen, A., Wang, J., Wang, L., Fan, J. W., & Liu, H. (2024). Explainable Diagnosis Prediction through Neuro-Symbolic Integration. ArXiv. https://arxiv.org/abs/2410.01855
18. Colelough, B. C., & Regli, W. (2025). Neuro-Symbolic AI in 2024: A Systematic Review. ArXiv. https://arxiv.org/abs/2501.05435
19. Pham, N. T., Kieu, T., Nguyen, D. M., Xuan, S. H., Duong-Trung, N., & Le-Phuoc, D. (2025). SLM-Bench: A Comprehensive Benchmark of Small Language Models on Environmental Impacts--Extended Version. ArXiv. https://arxiv.org/abs/2508.15478
20. Prieto, P., & Abad, P. (2025). Edge Deployment of Small Language Models, a comprehensive comparison of CPU, GPU and NPU backends. ArXiv. https://arxiv.org/abs/2511.22334
21. F. Pan et al., "Toward Software-Defined Vehicles: From Model-Based Engineering to Virtualization-Based Deployment," in IEEE Access, vol. 12, pp. 192127-192145, 2024, doi: 10.1109/ACCESS.2024.3512002.
22. S. Khokha, "Integrating Safety and Security in Autonomous Vehicles: A Comprehensive Review," 2024 4th International Conference on Sustainable Expert Systems (ICSES), Kaski, Nepal, 2024, pp. 460-464, doi: 10.1109/ICSES63445.2024.10763275.
