Every year, A2MAC1 purchases, scans, and tears down around 70 vehicles. Not as a demonstration, but as the foundation for some of the most detailed cost and engineering intelligence available in the automotive sector.
Sascha Voglgsang, Director of Costing and Insights at A2MAC1, joined Sasan Hashemi on the Beyond Cost podcast to explain what industrial-scale benchmarking actually requires, what the data reveals about EV cost structures today, and where the competitive gaps between legacy OEMs and Chinese players are widening.
The Real Meaning of Benchmarking at Industrial Scale
"Benchmarking" is a word used across industries for everything from a quick competitor review to a full engineering teardown. Sascha Voglgsang defines it with precision.
At A2MAC1, that definition is made operational at significant scale. The process starts before a single bolt is removed: a vehicle is selected based on market relevance and customer demand, purchased, and given a full visual 3D scan. From there, the teardown proceeds systematically, with the depth of analysis driven by where the engineering interest lies. For an EV from China, that means going all the way down to battery cell chemistry and semiconductor sourcing. For other vehicles, the priorities shift accordingly.
Why the Best-Resourced OEMs Still Rely on External Benchmarking
Most major OEMs do some form of benchmarking internally. The question is why so many of them rely on an external provider for the most comprehensive layer of it.
Sascha Voglgsang's answer is straightforward: quantity and quality.
No single OEM can maintain consistent teardown coverage across 70 vehicles a year, across multiple regions, with a standardized methodology applied throughout. Without that scale, any comparison between a Chinese newcomer and a European legacy manufacturer is incomplete at best.
Scale alone is not enough, though. A2MAC1's second advantage is the structure and granularity of the analysis. Going down to cell chemistry, raw material composition, and individual semiconductor manufacturer identification requires a depth of domain expertise that has been built over more than 25 years. Sascha Voglgsang describes the combination of coverage and granularity as the "winning recipe."
Building the Cost Layer
Technical benchmarking has been A2MAC1's foundation for over 25 years. The costing capability came later, built after the acquisition of a cost engineering consultancy. Sascha Voglgsang was part of that team. The goal was direct: for every vehicle analyzed technically, also evaluate the cost attached to each system and component.
In practice, this is more demanding than it sounds, because cost is not a straightforward measurement.
Consistency is one requirement. Scalability is the other. With a target of costing all 70 vehicles per year (currently at 18), manual expert review of every part is not viable. A2MAC1's approach is to automate what can be automated, thereby giving engineers the space to focus on genuinely complex cases, such as determining exactly how a giga-cast structural component was produced.
How A2MAC1 Is Using AI in Cost Engineering
Two concrete AI applications are already in use at A2MAC1.
The first is an inference agent that extracts part properties directly from teardown images. Rather than having a cost engineer manually review hundreds of photos to identify details like whether a part has a coating, the agent handles the identification. This feeds the bottom-up costing calculation with the right inputs, faster.
The second is the cost measure ideator. With a consistently structured database spanning hundreds of vehicles, the system can surface cost reduction ideas automatically. If a more cost-effective solution for a specific part category exists somewhere in the database, it can be identified and presented as a concrete measure.
AI, in this context, also functions as a quality control mechanism: every inconsistency in the data surfaces in the output, making it visible and fixable. But the methodology itself, including which design-to-cost levers apply to which part groups, still requires human governance. Subject matter experts define the rules; the system applies them at scale.
Three Patterns in the Data
A2MAC1's teardown database across hundreds of vehicles and multiple years makes visible patterns that are otherwise hard to quantify. Sascha Voglgsang highlights three.
1. Integration as a structural cost driver
Across batteries and power electronics, the direction is consistently toward fewer, more integrated components. Cell-to-pack, cell-to-chassis, consolidated power electronics modules: each step reduces housing weight, wiring complexity, and overall architecture cost. Legacy OEMs, facing EV market uncertainty in the early years, chose modular platforms that kept their options open. That decision carried a cost.
2. Supply chain localization, measured year by year
The BYD example is striking. In 2020, 2% of the semiconductors in a BYD onboard charger came from Chinese suppliers. By 2025, that figure is above 60%. The shift is deliberate, consistent, and carries a secondary benefit: localized supply chains tend to reduce carbon footprint alongside cost. As Sascha Voglgsang notes, European carbon footprint regulations are effectively functioning as a tariff in the same way. The two pressures are converging.
3. Shorter innovation cycles as a competitive strategy
While European OEMs have historically planned around long platform cycles, Chinese players iterate annually. Every BYD teardown since 2020 has revealed incremental improvements, architecture refinements, and further supply chain adjustments. The cumulative effect is a substantial and compounding cost advantage.
Looking ahead, software-defined vehicles are the next frontier. A2MAC1 can analyze the electrical architecture and identify which ECU controls which actuator, then assess the should-cost of software based on functional requirements. But as Sascha Voglgsang puts it: you cannot tear down software.
The data is available. The real question is whether your organization is able to turn it into a competitive advantage.
Listen to the Full Episode
Listen to the full episode to hear more about how industrial-scale benchmarking works in practice, how AI is beginning to reshape cost engineering at A2MAC1, and what the teardown data reveals about the competitive dynamics between Chinese OEMs and legacy manufacturers.
Watch the Full Video Episode
Listen on the Platform of Your Choice
