The whole play with making their own tooling seems ill conceived at Tesla's small scale, and it has definitely set the Model 3's production back quite a bit.
In fairness, Musk says they will eventually be as reliable, and as high quality as others and it'll be cheaper.. one day. That's the common NIH fallacy though.
The advantage of keeping much knowledge in-house is that you are not subject to extortion campaigns (e.g. the Prevent/VW dispute. In addition you
- are always exposing your internal knowledge to the part vendor and have to rely on the vendor protecting your IP from other customers, so "inventing here" works out better.
- are not bound to endless rounds of waterfall development. Tesla is heavily agile from what can be seen in reports... keeping communications (and decisions!) short saves money and especially time. Where time is Teslas biggest issue - just like a Tesla will kill almost anything at a drag race, eventually the gas-powered engine will win as it can go to higher speeds... and this is even more valid in terms of production. Tesla absolutely has to develop rapidly so they can get an established mass production e-car before Mercedes, BMW, VW and especially Toyota catch up and fly past Tesla.
- have to factor in margins for the vendors and by extension also their refinancing cost - Tesla gets shovelled with cheap cash, a vendor must build production lines with significantly more expensive capital. This adds up especially over many years...
- have less issues if something goes bad. For the old model, if a defect turns up it's usually a complex and long negotiation process where each side wants to avoid assuming blame (and thus costs). For Tesla, this cost risk falls away for all parts they manufacture in-house.
- have less probability for a catastrophic fuck up, I'd say. While a vendor has to "save face" and so has the incentive to e.g. hide issues that crop up during development or, worse, production from the customer (in order to avoid delay or other fines), the incentive for a Tesla-internal department is much less, as there are no delay fines.
... and thus very vulnerable to strikes or arbitrary price hikes. Yes, a seat is a seat, but once you're locked in into a vendor, you are locked in.
Tesla wants to avoid lock-ins as far as possible and especially something where a switch to a different manufacturer is as difficult as a seat. You can switch a vendor for a simple plastic part in a matter of days if the need arises (and the 3D data for the forming process is your IP), or do multi-vendor sourcing from the start - but try switching your seat vendor, that ain't gonna be cheap nor in a reasonably fast time frame.
> Yes, a seat is a seat, but once you're
> locked in into a vendor, you are locked in.
If it's easy for other car manufacturers at much bigger scales why would it be an issue for Tesla?
On the outside, it is a seat. What is different between manufacturers, and even between cars: dimensions of the components, abilities of the seat (e.g. how many degrees of movement, heating, massage functions), the attachment of the seat to the car itself, the form of the seat... if you want to switch the seat vendor, you have to account for weeks if not months until the new vendor delivers the same final product.
> If it's easy for other car manufacturers at much bigger scales why would it be an issue for Tesla?
They don't switch or only switch at model "refreshments", which is at odds with Teslas always-improving workflow.
If they have professional supply chain and commercial departments they'll have contingencies in place for situations like this.
Even companies as massive as VW can not handle such a situation (as evidenced last year: https://www.ft.com/content/ffd3ee62-6934-11e6-a0b1-d87a9fea0...).
And they've brought in house more since then (like the seats).