Part I of this series examined how large-scale platform modularity, particularly in EVs, can introduce design tradeoffs and system-level ripple effects once real engineering constraints appear. Promised flexibility often gives way to added complexity, compromise, and systems that are harder to design, build, and support than anticipated. Part II turns to those concerns directly, examining where modularity works and where it genuinely delivers value.
Not always bad, but acknowledge realities
These concerns do not mean that modular platforms are always a bad idea; that’s not the case at all. For example, if you are upgrading to a faster processor or one with more memory but with the same footprint and pinout, that’s probably a good idea (just watch the thermal impact).
Or consider a classic “mainframe” oscilloscope such as the renowned 1-GHz Tektronix 7104 from 1979, with four uncommitted plug-in bays in its lower half, visible in Figure 1. These can be populated as needed with different input amplifiers, spectrum analyzers, and other specialty plug-in modules, all with no negative consequences to the basic operation. It’s a platform for modularity at its best.
There’s ample precedent for wondering and worrying about the overall impact on design, manufacturing, test, and in-use support when you attempt to modularize and build up in a logical manner, admittedly from a different context. My perspective is from the book “Apollo: The Race to the Moon” by Charles Murray and Catherine Bly Cox, a most insightful book on the moon landing.
In a brief overview chapter, the authors point out that modular, stage-by-stage assembly of the rocket and capsule didn’t accomplish what people thought it did. The book notes that NASA project leaders realized this after a successful first-stage test, noting that “whenever you added a new stage, the ground support equipment was different, the checkout procedures were different, the hardware was different.”
The alternative strategy the Apollo project had to use for assembly and test was non-intuitive and called “all-up.” It meant that from the start, they would assemble the rocket in its entirety and only in its final configuration, using subassemblies which had been individually tested but not joined to others yet.
They would then test the integrated rocket as a single completed system. This is in contrast to building it up piece by piece and testing at each incremental build step, which at first seems the more logical way to do it.
Surely, the engineers and others at Stellantis have studied this idea very carefully before implementing it. It will certainly be interesting to see how manufacturability, performance, support, and costs work out. Still, one wonders how much of the modularity story is based on actual savings and how much is driven by the conceptual attractiveness of the story (especially to Wall Street), more than the reality. We’ll have to wait and see about that.
References
- Stellantis Reveals STLA Medium Platform Designed to Electrify the Heart of the Global Market with Future-Proof Customer Innovations, Stellantis
- Stellantis Unveils BEV-native STLA Large Platform with 800 Km/500 Mile Range and the Ultimate Flexibility to Cover a Wide Spectrum of Vehicles, Stellantis
- Stellantis STLA Medium Platform, Stellantis
- Why Stellantis’s CEO Remains All-In on EVs as Others Retrench, The Wall Street Journal
Related EE World content
- Platform helps build IoT networks
- Microtune introduces new chip platform
- Harmonic introduces DWDM transmission platform
- Intelligent High-Power Modular Platform
- Modular Concept Based on Robust Components
- Modular system combines DMM with DAQ
Filed Under: EE World - EV ENGINEERING, FAQs, Featured Contributions