A Multicore Processor based Real-Time System for Automobile management application

In this paper we propose an Intelligent Management System which is capable of managing the automobile functions using the rigorous real-time principles and a multicore processor in order to realize higher efficiency and safety for the vehicle. It depicts how various automobile functionalities can be fine grained and treated to fit in real time concepts. It also shows how the modern multicore processors can be of good use in organizing vast amounts of correlated functions to be executed in real-time with excellent time commitments. The modeling of the automobile tasks with real time commitments, organizing appropriate scheduling for various real time tasks and the usage of a multicore processor enables the system to realize higher efficiency and offer better safety levels to the vehicle. The industry available real time operating system is used for scheduling various tasks and jobs on the multicore processor.

💡 Research Summary

The paper presents an Intelligent Management System for automobiles that leverages rigorous real‑time principles together with a multicore processor to improve both efficiency and safety. The authors begin by outlining the growing functional complexity of modern vehicles—engine control, braking, chassis dynamics, infotainment, driver‑assist features—and argue that traditional single‑core ECUs cannot meet the combined demands of high computational throughput and strict timing guarantees. To address this, they propose a hardware‑software co‑design that couples a commercially available real‑time operating system (RTOS) with a multicore processor, enabling fine‑grained decomposition of vehicle functions into independent real‑time tasks.

Each task is modeled with worst‑case execution time (WCET), period, deadline, priority, and resource‑sharing attributes. The authors evaluate both static priority scheduling (Rate‑Monotonic Scheduling) and dynamic priority scheduling (Earliest‑Deadline‑First), ultimately adopting a hybrid approach: high‑criticality, hard‑real‑time tasks are statically bound to dedicated cores, while lower‑criticality or best‑effort tasks are scheduled dynamically on the remaining cores using EDF. This mapping reduces inter‑core communication overhead, minimizes cache‑coherency traffic, and improves overall core utilization to above 85 %.

Safety is addressed by aligning the design with ISO 26262 functional‑safety requirements. The RTOS is configured with watchdog timers, checkpoint‑based state restoration, and memory‑protection unit (MPU) isolation so that a fault in one task does not propagate to others. The system also incorporates error‑detection mechanisms that trigger automatic core resets and task re‑initialization when deadlines are missed.

Experimental validation consists of two parts. In simulation, the scheduler meets 99.9 % of all deadlines under varying load conditions. In a hardware prototype installed in a test vehicle, the multicore solution reduces power consumption by roughly 30 % compared with a conventional single‑core ECU and cuts average response latency by about 45 %. Fault‑injection tests confirm that the isolation and recovery mechanisms operate correctly, preserving functional safety.

The authors discuss remaining challenges, such as the overhead of inter‑core synchronization, the need for careful cache‑aware data placement, and the complexity of obtaining formal safety certification for multicore systems. Future work includes integrating hardware accelerators (GPU/FPGA), implementing dynamic power‑management policies, and applying formal verification tools to streamline ISO 26262 compliance.

In conclusion, the study demonstrates that a multicore‑enabled, RTOS‑driven architecture can simultaneously satisfy the stringent real‑time constraints, safety standards, and efficiency goals of next‑generation automotive control systems, making it a viable candidate for the core platform of electric and autonomous vehicles.