EV Engineering & Infrastructure

Part I: Why it’s important to be wary of “platform” and “modular” architectures

By |

Editor’s note: This article is particularly relevant to electric vehicles because EVs are far more architecture-sensitive than ICE vehicles. Changes in voltage level, battery chemistry, drive layout, or regenerative braking ripple through the chassis, electrical system, thermal management, and control software, making the limits of platform modularity more visible in EV applications.

The term “modular” usually has positive connotations. It can refer to something as basic as a chassis into which you plug various test or I/O cards, or as complicated as a house that’s built using individual room designs built in a factory, which are then trucked to the building site and assembled into a complete house. It’s also a word which often conceals — either intentionally or inadvertently — what is involved in a project.

The presumed benefits of modularity have been embraced over the years by automakers, among others, who use product-line “platforms” on which they build somewhat different vehicles. In nearly all these situations, the modularity is mostly at the “trim” level and much less so in the power train, by using a single or closely related engine-transmission pairing, but with different body styles, interior styles, and finishes.

But why stop there? I was intrigued when I saw a story about a plan by Stellantis NV to use a highly modular automotive platform for next-generation vehicles. Dubbed “STLA Large,” this platform can accommodate many power-train options, shown in Figure 1.

Figure 1. The drawing of the bare platform gives no indication of the end product for which it is intended. (Image: Stellantis)

Note: Stellantis is the parent company of Chrysler, Peugeot, Fiat, and many other well-known nameplates; their CEO at the time of this ambitious announcement, Carlos Tavares, has since been fired as his grand visions and their execution were counterpunched by reality.

The proposed flexibility covers propulsion — front-drive, rear-drive, all-wheel drive, and multi-energy, which are provided by a family of three, scalable electric drive modules.

Wow, that’s a lot of modularity and flexibility. This large platform is one of four different-sized ones they have developed, primarily but not exclusively for battery-electric vehicles (BEVs).

According to their January 2024 press release, these four modular STLA platforms  (Small/Medium/Large/Frame) are engineered to be “future proof” (always a worrisome cliché) and are inherently flexible in wheelbase, width, overhang, ride height, and suspension design.

Platform modularity

They’re designed with provisions for future battery chemistries, including nickel and cobalt-free and solid-state batteries.

STLA Large is designed and engineered as a native BEV platform with the option of 400 and 800-volt electric architectures, as seen in Figure 2. It can be configured in front-wheel-drive, rear-wheel-drive, and all-wheel-drive layouts using transverse and longitudinal engine mounting configurations.

Figure 2. The STLA Large is a native BEV platform and can accommodate 400 and 800-volt electric architectures. (Image: Stellantis)

In addition to BEVs, the STLA Large platform also supports hybrid and internal combustion engine (ICE) propulsion systems, as shown in Figure 3, while the smaller ones are BEV only.

All this modularity and associated flexibility seems like a very sensible idea. It may even have impressed stock analysts, although they since gone negative on Stellantis for many reasons.

Figure 3. The STLA Large can also support hybrid and internal combustion engine power trains. (Image: Stellantis)

But (and engineers should always think about that “but”), as both a product designer and a user, in many cases, the benefits and results were less than anticipated.

Primarily for two reasons:

First, there’s the issue of design optimization. Any individual product design requires tradeoffs and compromises to balance competing demands with respect to weight, space, power, cost, safety, components availability, and test. and many other factors. It’s hard enough to accomplish this when you are focused on a single design.

Doing so for a modular multi-use platform requires significant additional compromise on features and performance to accommodate the various permutations and arrangements within a common framework. Sometimes these compromises are minor or modest, but often they are not. You end up with a severely sub-optimum design that sounded like a good idea at the beginning.

Instead of a “very good” non-modular design, you end up with a “sort-of good” modular one. The size of the gap between very good and sort-of good is hard to know without careful analysis. Designers may have to do complicated “gymnastics” to make the modular platform concept work at all, let alone well.

Second, there’s the belief that “modular” means you can just drop in “unit B” to replace “unit A” and do so with minimal additional changes to the platform, interconnections, and supporting peripheral functions.

Once again, experience has shown that there are countless other less-visible changes you need to take to fully accommodate the drop-in alternative. The ripple effect of switching out even a single modest component can bring nasty surprises.

For example, weight and weight distribution will likely be different, so suspension will change; different sensors are needed for each type of power train; and cabling, wire gauges, and connectors will be very different as well. A detailed review of the manufacturing bill of materials (BOM) and process will reveal many other hardware, inventory, and production-process changes in addition to the supposedly “painless” one of different software packages.

I’d guess that at least 80% of each modular platform version will be different from the others. In short, the likely pain may be far greater than the actual gain, and the hoped-for benefits will be illusory.

Think about something as distinctly modular as the regulative braking subsystem in an EV or BEV but not offered in an ICE vehicle. There will be the regenerative-braking hardware at the wheel, of course, as well as cabling, interconnects, and system-level software.

In reality, even this distinct power-harvesting subsystem is not a simple drop-in add-on, yet it is only one modest difference between vehicles on the same platform.

Part II looks at some bigger concerns as well as where modularity and platforms make technical sense.

References

Related EE World content

Filed Under: FAQs
Tagged With: , , ,
 

 

Search EV Engineering & Infrastructure