Software Defined Vehicles
Interview with Cedric Armand, Ford
“ADAS is the primary domain to benefit from virtualization at scale”
As SDVs push software complexity and validation demands to new levels, development speed becomes a critical bottleneck. Ford’s Cedric Armand explains how virtualization reshapes software delivery, testing and system integration.
As software-defined vehicles increase system complexity across domains, traditional development and validation approaches are reaching their limits. Sequential ECU workflows, late-stage integration and hardware dependencies are becoming key bottlenecks for speed and scalability.
Cedric Armand, Director of Virtualization at Ford Motor Company, focuses on how virtual platforms can fundamentally change this paradigm. Leading global efforts around virtual ECUs, cloud-based development environments and software validation strategies, he works on enabling earlier integration, faster iteration and more resilient development processes.
At the Automotive Computing Conference 2026, Armand presented how virtualization accelerates automotive compute by shifting development and validation into scalable, software-driven environments. In the following interview, he explains where virtualization already delivers measurable gains and what organisational changes are required to unlock its full potential.
ADT: What role does virtualization play in managing growing software and system complexity?
Advertisement
Armand: Virtualization addresses the exponential growth in system complexity, driven by the intricate interplay between software and physical components, by providing a high-fidelity, repeatable environment that physical hardware cannot consistently replicate. As the permutations of these interactions increase exponentially, virtualization mitigates the resulting “loop-to-loop” variation inherent in physical systems, allowing for truly repeatable scenario testing and significantly accelerated development cycles. This capability is essential for exploring vast design spaces and identifying critical “corner cases” through techniques such as Monte Carlo simulations, which can be executed efficiently and without the risk of hardware damage. Ultimately, virtualization transforms an otherwise unmanageable design space into a structured, probeable environment, ensuring robust performance across the entire automotive computing stack.
How do virtualized environments change validation cycles and development speed compared to traditional ECU workflows?
Virtualized environments fundamentally transform traditional, sequential ECU workflows into a parallelised “shift-left” development model by decoupling software maturity from hardware availability. By utilising high-fidelity virtual prototypes, engineering teams can initiate software development and validation months before physical silicon or printed circuit boards are available, effectively front-loading the development timeline and mitigating the high-risk “integration hell” typical of late-stage hardware-in-the-loop testing. This early-stage visibility allows for the identification and remediation of complex bugs when they are least expensive to fix and least disruptive to the programme, creating a more resilient path towards production. Ultimately, by maximising the use of software simulation and virtual prototyping, OEMs can achieve a significant competitive advantage, with industry benchmarks demonstrating a reduction in time-to-market of nine to eleven months.
What impact does virtualization have on cross-domain collaboration between software, hardware and safety teams?
Virtualization acts as a bridge between historically siloed domains, software, hardware and safety, by enabling a “shift-left” approach to system integration. Traditionally, cross-domain collaboration is delayed until physical system integration milestones, where the discovery of inter-system conflicts often causes costly setbacks. Virtualization disrupts this cycle by allowing teams to integrate virtual controllers and high-fidelity models of external subsystems early in the development process, creating a shared digital environment for holistic validation. This allows domain experts to simulate and refine complex inter-system interactions without the logistical burden of sourcing scarce prototype hardware or the technical overhead of maintaining physical systems outside their area of expertise. Ultimately, virtualization fosters a more proactive, integrated engineering culture, ensuring that safety-critical dependencies and software-hardware interfaces are harmonised long before the first physical vehicle is assembled.
What organisational or cultural barriers still slow down adoption of virtualization in automotive development?
The primary organisational barrier to the adoption of virtualization is the persistence of departmental silos, which fundamentally conflict with the “shift-left” philosophy. Because virtualization requires critical integration points to occur much earlier in the development lifecycle, software developers, hardware engineers and validation teams can no longer operate as sequential, isolated units. Overcoming this requires the implementation of integrated workflows supported by a “digital thread”, a unified data platform that ensures all stakeholders operate from a single source of truth. By breaking down these structural barriers, organisations can enable the real-time collaboration and cross-functional decision-making necessary to manage the interdependencies of modern, software-defined vehicle architectures. Culturally, the transition is often hindered by a deeply ingrained hardware-centric mindset that views integration as a late-stage, reactive milestone rather than a proactive development driver. To move forward, organisations must adopt a philosophy of “testing early and often”, prioritising the reduction of feedback loops and the continuous improvement of automated, scalable testing practices. By establishing this proactive approach as a cultural “North Star”, OEMs can shift their focus from resource-heavy physical validation to efficient, simulation-led development. This mindset shift ensures that integration testing becomes a foundational element of the creative process, allowing the right technological innovations to follow naturally from a more agile and resilient development framework.
Where does virtualization already save months in development cycles today beyond just reducing complexity?
Advertisement
Virtualization delivers substantial time savings by fundamentally de-risking the integration phase, traditionally the most volatile and time-consuming period in automotive development. By enabling high-fidelity validation early in the lifecycle, 65 per cent of defects that would normally be discovered during integration are identified and resolved before they ever reach a physical system-level bench. This “defect containment” is critical because root-causing an issue during the late-stage integration phase is up to ten times slower and more resource-intensive than addressing it in a less complex, virtualised environment or component bench. By shifting these diagnostic efforts to an earlier, more efficient window, virtualization eliminates the compounding troubleshooting delays that typically derail programme timelines, resulting in a significantly more streamlined and predictable path to market.
Which vehicle domains benefit first from virtualization at scale, ADAS, IVI, body or powertrain, and why?
ADAS is the primary domain to benefit from virtualization at scale, driven by the dual imperatives of logistical feasibility and safety-critical validation. The requirement to verify millions of probabilistic behavioural scenarios and edge cases makes physical testing fundamentally impossible within reasonable development timelines. Moreover, virtualization provides a necessary, risk-free environment to test high-stakes manoeuvres, such as pedestrian interactions at intersections, which cannot be safely performed in the physical world until the system has reached a high level of maturity. Following ADAS, the in-vehicle infotainment domain realises significant benefits due to the complexity of modern software stacks and the wide variety of user interaction scenarios. By virtualising these domains, OEMs can probe interaction spaces that would otherwise remain only partially tested, supporting both functional safety and user experience.
Partnerships are becoming essential across the automotive computing stack. Which type of partnerships will matter most in the next three years, strategic, technological or regulatory, and what should OEMs prioritise first?
OEMs must prioritise strategic partnerships as the primary catalyst for transformation over the next three years, as they establish the foundational architecture for both the business model and the technology stack. This prioritisation begins with a definitive “build versus buy” strategy, where the OEM identifies the specific layers of the computing stack that deliver unique, brand-differentiating value to the end customer while treating non-differentiating components as commodities. By establishing these strategic frameworks first, OEMs create a financially sustainable and scalable environment that allows technological progress to move at the speed of the software industry rather than the traditional hardware cycle. Ultimately, these high-level alliances provide the structural clarity needed to navigate immense capital requirements, ensuring that every subsequent technological investment is both purposeful and economically viable.
