Modern vehicles rank among the most sophisticated software systems ever developed. At the same time, market expectations continue to accelerate, forcing automakers to compress development timelines.
To meet this challenge, the most advanced players are already weaving artificial intelligence into their engineering processes. The aim is to shrink diagnostic cycles that once stretched across weeks to just a few hours, while also enhancing forecast capabilities.
By leveraging data from development and testing, AI helps teams spot anomalies more quickly, understand their root causes, and fix issues before production begins.
In this context, time-to-market now reflects a manufacturer’s ability to continuously evolve its vehicles, rapidly address field defects, and integrate new features throughout the vehicle’s lifecycle. This agility is becoming a real differentiator.
Vehicle Diagnostics: A New Lever for Acceleration
Diagnosing a fault now requires understanding the interplay between hardware components, embedded software, sensor data, maintenance histories, and real-world usage conditions. Traditional, often reactive approaches struggle to keep up with this complexity.
AI thus enables quicker identification of probable causes, prioritization of corrective actions, and helps both engineers and technicians avoid pursuing dead ends.
For development teams, during preproduction phases when systems are still incomplete and faults are hard to isolate, AI can substantially cut analysis time and speed the transition to mass production start-up (Start of Production – SOP).
This issue is especially important given that a preproduction vehicle can cost up to ten times more than a high-volume production vehicle due to the absence of industrial automation and scale economies. Every week gained in testing cycles has a direct impact on budgets, while lowering the number of prototypes required amplifies these savings.
Concretely, AI-assisted automatic sorting of event logs can cut the analysis time by up to 80%. Moreover, preprocessing data directly in the vehicle can reduce cloud processing costs by 60%.
These gains translate into faster development cycles and lower engineering costs.
Maintenance Becomes an Economic Lever
Waiting years to adjust a feature, fix a software behavior, or improve the user experience is now increasingly unsustainable. The vehicle is evolving into a scalable platform capable of progressing throughout its lifecycle.
A recent study on the evolution of software-defined vehicles (Software-Defined Vehicles/SDV) confirms this trend. On a global scale, predictive and proactive maintenance has become in 2026 the leading driver of customer loyalty and aftersales revenue, cited by 44% of respondents. In France, the figure reaches 47%.
This shift comes at a pivotal moment for the industry. Today, only 9% of vehicles have reached SDV maturity at level 4, while 77% of sector professionals anticipate wide SDV deployment by 2030.
In other words, the gap between ambition and implementation is rapidly narrowing.
With Data, Time-to-Market Becomes Continuous
Accelerating development up to the point of market launch is only the first step. Value creation can continue long after the vehicle hits the road, through software updates, feature enhancements, and corrections of real-world behaviors.
Moreover, the data generated by vehicles in service feeds continuous-improvement loops. An issue detected on a subset of the fleet can help refine future models, tighten software, or optimize maintenance strategies.
Chinese automakers are already illustrating this new approach by using fleet data to roll out certain features on a weekly basis rather than once a year.
France, however, maintains a relatively distinctive approach. While the global trend leans toward internal data usage (enhancing diagnostics, boosting performance, or developing new services), 52% of French respondents still view data monetization—through applications such as usage-based insurance or V2X services—as a high-value opportunity.
This ambition is valid, but it should not obscure the need to orchestrate, contextualize, and effectively exploit data to improve the vehicles themselves before monetizing it.
Thus, AI can become the tool that unites all stages of time-to-market, provided it is designed as an engineering infrastructure, not merely as a technological add-on.
*Alexandre Corjon is Senior Vice President of Engineering at Sonatus
