Software-Defined Vehicles: Transforming Automotive Engineering Through Modular Architecture And Ai-Enabled Validation

Abstract

The research of this paper aims at discussing the prospects of the change of the world of automotive engineering with the help of Software-Defined Vehicles (SDVs) with the help of modular architecture, AI-assisted validation and model-based systems engineering (MBSE). SDVs will formal detection digitalized divide by centralized, zonal plan which will reduce wiring by up to 45 per cent using 2-dose of the mean duration between failures that typical structures. It has hit speeds of 190 MB/s with Cellular Vehicle-to-Everything (C-V2X) and 5G transmission rate and thus complete software updates with OTA which is reportedly 70 times quicker than systems controlled via Wi-Fi and 5000 times more adept than software update. Comparison between AI based validation and scripted coverage gave 81-percent bad code coverage compared to 62-percent scripted coverage at half the time. The requirements traceability score has achieved 25 points out of the required 78 to 97percent and preparations in certifications have reduced by 40 percent because of MBSE integration.

This means that SDVs do not have to be commodities, but instead can be constructed and delivered as running platforms. The findings also indicate that AI and MBSE also minimize engineering labour and attain a safety, reliability, user-trust increase. Study revealed that it required incorporation of modular architecture and intelligent testing, among others, so as to keep abreast of the increasing complexity of the automotive systems. Among its implications are the decrease in the cost of maintenance, the steepening of the timeline of software adoption, the growth of the level of trust in autonomous or semi-autonomous vehicles among the population.