Richard
Wang.

Frontal area is said to be a key factor influencing a car's aerodynamics. Sometimes there are outliers — e.g. BMW M4 GT3 — but in general we assume a car with less frontal area will experience less drag if we leave everything else equal. For instance, if we sawed off the A-pillars of a Ferrari 296 GTB, lowered the enclosed rooftop, and let the driver take up both seats lying sideways, then compared it to a regular 296 GTB: the former should go faster. I believe most people would agree on that.

Venturi ducts are a debatable feature when it comes to improving aerodynamics. In many cases they do help increase downforce, depending on placement. However, whether they help reduce drag despite reducing frontal area is uncertain. The answer is more likely negative, but it also depends on the structure.

The 1967 OSI Silver Fox was built with a huge center duct to break the 24hr Le Mans record, but never got effectively tested because the company went bankrupt. The performance of the WMC250EV — also with a center duct — seems to justify the concept.

S-ducts are more common these days, especially on hypercars. While the Lotus Evija stands out as one of the most aggressive, it had lower stats (top speed, Cd value, etc.) compared to other hypercars around the same era — despite tremendous downforce and 2,039 hp.

The conclusion: while the pressure difference created by s-ducts may increase downforce but worsen drag tolerance, a large enough center duct might have been beneficial for reducing drag and achieving higher top speeds. There is a lot of work in terms of converting the structure — the main reason we don't see modern quads with this technology.

Inspiration behind EPCX

There were two (debatable) automotive features I was looking into before I ever wanted to build an automobile, both in high school. One was adding a HALO to a road-legal go-kart; the second was prone seating for street-legal 4-wheelers.

I guess a fair reason for some people applying to Princeton was that the Princeton Racing Electric had overall better aesthetics compared to many other Formula SAE cars. It might be more fun to modify the body and add a HALO, to make the "not-so-street-legal" car street-legal. Apparently, BAC Mono and KTM E-Bow do a better job at this for a production version. The following was a reimagined rough sketch from what I submitted along with my application in 2021:

A car with a kneeled seating structure — where three prone seats are placed the same way as the McLaren F1/Speedtail, with one additional seat behind the driver's seat — doesn't sound very smart. But I did submit an annotated sketch (now lost) along with my application to the University of Chicago back in Jan 2021. Hennessey's Project Deep Space had a roughly similar layout but with regular seats; now seems to have gone south.

The most widely recognized procedure for developing automotive concepts is to start from sketches. Some rare examples in history use scale models only — no sketches at all — and eventually get to a fully-functional prototype. E.g. Lotec Sirius.

As an extreme scale-model person with minimal professional sketching experience, I knew sketching wasn't the path for me if I were to execute it.

My proposed approach

• Build-from-scratch a physical model
• Scan it into a digital 3D model
• Adjust for size & mirroring (if necessary)
• 3D print & hands-on modifications
• Have specialists render, engineer, and test on it
• Repeat 1–5 until maximum satisfactory

A very untraditional way for Part 1 is to start from an existing bodywork that has the overall best-fit structure. A very extreme example is the conversion from the Ferrari 488 GTB to a Touring Superleggera Arese RH-95 (the real ones were also converted from 488 donor cars):

This technique still outperforms advanced prototyping on the DIY market by a few indicators, but is highly dependent on personal skills. We expect that scaled 3D Gaussian Splatting will soon make parts of this workflow smoother and more precise than ever before. A few friends of mine in automotive design have been helping me refine along the way. We hope for some executions in the near future.

At some point, Koenigsegg decided they would convert a Regera into a camera car to shoot their record-breaking moments, meaning the "Huracam" is no longer the fastest camera car in the world. While the extra equipment attached to the Regera may look very inconvenient and would cause some unwanted turbulence, perhaps "enough horses is all you need." The exact same concept was later well-played by the team behind the Yangwang U9 Xtreme. Let's attempt a more aerodynamic and better-looking camera system for the Regera, and call it both the fastest and the most aerodynamic camera car.

The KACS unit 1 is basically an aerodynamic frame fitted with a single robotic arm where a PTZ camera can be hung underneath (usually smaller ones). The frame comes with a specially designed targa top and extends over the top of the engine cover, similar to the looks of the One:1 spoiler. The maximum lens span can extend to 270 degrees — but since the forward vision is blocked by the car body, the system can only be used to shoot things sideways, mostly.

To maximize lens span, we can sacrifice some aerodynamics by mounting the camera on top of the frame instead of underneath, so no structure blocks the way. To achieve the highest aerodynamic efficiency, the KACS unit 2 also comes with an all-new curved robotic arm that is tuned to self-adjust its structure based on real-time car movement, while making sure it doesn't block the lenses. Now we get a camera car that is not only faster than a Huracam, but can also shoot blind-spot-free with less arm span compared to a traditional U-crane. We'll call this the KACS unit 2.

"Automation" has become one of the default words of the Gen Z founder circle, especially now that AI workflows are making it easier to automate, optimize, scale, and remove friction. It is, however, not always what really makes people feel alive — or what makes the known world more interesting for the vast majority.

Thinking beyond automation

What is a product so good that people feel irrational urgency to own it?

What is a technology so strange, beautiful, or powerful that it changes the benchmark for what people think is possible — think of the CG used in Transformers?

For a long time, "copying from China" was treated as a lazy idea since a common assumption for the past century was that China was good at execution, scale, and speed, instead of original creation. It's amazing how that assumption is increasingly outdated. It might be the time to copy back.

Over the past few decades, China has become one of the most intense laboratories for new consumer behaviors, distribution models, offline-online systems, and urban-scale products. While part of this came from people studying abroad and bringing ideas back, a lot more of it came from something less appreciated:

• Massive local experimentation under very different constraints;
• Consumer economy favors rapid experimentation & scale straight from the individual level, as the society's structure doesn't support great creativity at a corporate scale, something even many Chinese locals underestimate.

We are seeing many ideas that seem strange or impossible in the U.S. already existing in China as working products, physical spaces, service systems, or entire industries. They were not always visible to the West due to being embedded in a different language, platform ecosystem, city structure, and consumer rhythm. At the same time, China is no longer only "absorbing" ideas from abroad. Increasingly, ideas born or stress-tested in China are beginning to influence the rest of the world — sometimes directly, sometimes through quiet imitation, and sometimes through founders who understand both systems well enough to translate one into the other.

"Are we patient enough to listen to these immigrant / international founders that seem odd" becomes the more important question beyond "how do we find these founders": building on top of these market gaps involves understanding the hidden mechanism behind the invention, then rebuilding it with U.S.-native assumptions, adjusted distribution models, and different taste. This will not be obvious work. While immigrants and third-culture founders continue to have the natural advantages, the ways they conduct and apply immense real-world comparison — i.e. comparing entire ways of living, consuming, trusting, moving, paying, and coordinating — is a rare skill native observers struggle to understand and apply.