Software is at the heart of modern mobility. From advanced driver-assistance systems (ADAS) to electric vehicle platforms, it now defines performance, safety, and innovation in the automotive industry.
Yet with this opportunity comes risk: the number of software-related recalls has risen sharply in recent years, with software fixes now accounting for more than one in five recalls. These are not minor glitches; they often involve ADAS features such as automatic emergency braking, powertrain controls, and critical electrical systems.
Go deeper with GlobalData
Discover B2B Marketing That Performs
Combine business intelligence and editorial excellence to reach engaged professionals across 36 leading media platforms.
This growing dependency on software introduces a fundamental challenge: How can manufacturers ensure reliability and safety when systems’ complexity expands at a pace that traditional methods struggle to keep up with?
ISO 26262: More than compliance
The ISO 26262 standard, which governs functional safety for electrical and electronic systems in vehicles, has become the reference point for managing this challenge. Compliance is not simply a technical checkbox; it is a strategic necessity.
Yet achieving this level of assurance is not trivial. Testing and static analysis alone cannot provide complete confidence when faced with millions of potential inputs and interactions.
This is where formal methods, and in particular, technologies like SPARK, are changing the equation.
The case for formal methods
Formal methods apply mathematical proofs to software development, offering guarantees that go beyond conventional testing. Instead of extrapolating correctness from a limited number of test cases, formal verification demonstrates that critical properties hold true for all possible inputs.
This capability is especially relevant for higher Automotive Safety Integrity Levels (ASIL C and D), where the consequences of failure are unacceptable. Formal methods can prove:
- Absence of run-time errors
- Conformance to functional requirements
- Secure information flows within the system
In practice, this means that software underpinning emergency braking, steering assistance, or battery management can be developed with mathematical certainty around its safety properties. Something conventional approaches cannot deliver on their own.
The SPARK language and tool suite are particularly well-suited for automotive software development, aligning with the rigorous demands of ISO 26262. By enabling static verification and formal proof of correctness, SPARK eliminates entire classes of run-time errors, such as buffer overflows and data races, which are critical concerns in safety-critical automotive applications.
From concept to reality
Advances in proof technology and supporting tools have made formal verification scalable and practical for production environments.
As first reported by Just Auto in June 2025, organisations like NVIDIA have already integrated SPARK into their development processes, achieving the highest levels of assurance for complex, safety-critical applications.
