I definitely have a basis for Specialized. I base this off snippets in a few articles here and there (that mention FEA, CFD and their wind tunnel), job board postings for engineers at specialized, and a look at their LinkedIn profile showing their employees (background and training).
I have a dislike for a number of brands because when I do hear about their design team, it’s just one or two unqualified people who essentially guess. (Note, I’m an engineer who deals in design and manufacturing of medical devices. I actually really enjoy this stuff. So I’m super biased to testing, process controls. Less emotional/feeling based design … “It looks fast and strong” … I’m more “testing a sample size of n has shown that we have 99% confidence that 99.5% of the frames are not going to fail when ridden in 99.9% of the real world conditions” … more that sort of thing.
I would love to read a summary of some of the times Specialized got it wrong. Remember the SL2 Tarmac had a fork recall? Who got it wrong then? Was it a type 1 issue where they designed it wrong and somehow their R&D testing didnt catch it? Or was it designed right, but some process variation wasn’t noticed and they made some that were too light and didnt have enough meat in the right spots?
I bet someone will read this and think “no, Bobby is a custom frame builder who’s been making beautiful frames for 40 years and never had one break. He’s a real artist”. From my engineering point of view, Bobby is unlikely to be making the frame on the lower limit of weight, because he can’t know how overbuilt it is. And equally if he released a new model and it was right on the limit, so maybe one in 10 frames will fail when they hit a speed bump at speed … if he only built 5 prototypes, and it’s a 1 in 10 failure … how would he know?
In truth, most companies probably overbuild their frames. As I look at higher and higher end bikes that are built lighter and lighter, then I want to know they’re based on methodical design/development, and rock solid process controls. If 850 grams is where the SL7 S-works is (I’m guessing), then once you get up to 1250 grams, then I think the margin of error is wider and I’m not as worried. But for sure, I’m not buying a no-name Aliexpress knockoff of and SL7 that weighs 800, 900 or even 1000 grams.