Gales of change: the simultaneous revolutions in the auto-sector Part I.

20 years ago automakers were stuck on an idea that future electric cars would have “skateboard” construction: with a flat base, bodies that could be swapped out depending on your need/mood, batteries that could be swapped out depending on charge and electric motors in each wheel hub for packaging.

10 years Tesla launched their Model S, the first successful mass-market electric vehicle that crystallised the forms and paradigms for successful electric vehicles to come - no wheel motors, nor swappable bodies/batteries, but importantly front trunks and giant centre screens.

This year was the first time an electric vehicle, the Tesla Model Y, was the highest selling car in the world, pipping the Toyota Corolla.

In my last couple of pieces I looked at the electric vehicle (EV) transition from a demand side/consumer perspective. Now I want to share my personal take on the EV transition from the supply side/automaker perspective. Making predictions is hard, especially about the future. Tesla has gotten a lot of this right, and in many instances it’s a case of the legacy automakers playing catch up to where Tesla was 7 years ago, while Tesla charges ahead.

There has been some general commentary on Tesla's vertical integration but that misses a lot:

1. Any nuances including Tesla’s battery partnerships with Panasonic
2. Competitors dabble in deeper vertical integration, e.g. Hyundai have smelted, rolled, and pressed their own steel for decades
3. the general fact that vertical integration is a means to an end.

The industry and sector is changing rapidly - 2023 was first time the USA hit 1 million EV sold within a calendar year. Globally about 13% of new passenger vehicles sold are new energy. Digging a little deeper than the electric cars sales, what you find is a sector where everything seems to be in turmoil everywhere, all at once.

There has been one major bit de-risking for the sector in 2023. Biden’s Inflation Reduction Act, has sparked the industry coalescing on to some charging (read: plug) standards, so consumers will no longer be forced to pick a winner. But beyond that, let me run through the disruptive mega-trends of chaos the vehicle makers are currently navigating. This will be in two parts, starting with the all the disruptions instigated by Tesla.

Electric propulsion

This is the obvious one to nod at quickly. The electric vehicle revolution is an electric propulsion revolution.

?The technical competencies required for an electric powertrain are very different, as is the supply chain. Moreover, there are fundamental questions to be answered - for each of the major components there’s an open question about whether the car makers want to do it themselves or buy from a supplier and this is changing rapidly .?

There are two common misconceptions that come with electric propulsion: innovation and containment.

Innovation

Innovation isn't just occurring in battery technology and infotainment. Even though electric motors are something that Benjamin Franklin tinkered with, and have been commercialised for over a century there are still big improvements being made in recent times. An example is the revolutionary ‘hairpin’ electric motor design, first released by GM in 2008, only now is Tesla committing to the technology.

Much focus on innovation across the EV power train is on production costs (both methods and materials) and efficiency rather than performance (think power and torque output). We are already hitting our head on the ceiling when it comes to appropriate amounts of power to sell to the general public.

Kia offer their EV6 family car in a trim level that accelerates from 0 - 100 km/h in 3.5 seconds - making the car more accelerationery than a top-of-the-line 2009 Lamborghini Murcielago LP 670-4 SV or 2006 Porsche Carrera GT. I predict many cars will have software limits in Australia so they don’t hit requirements of the new ultra high powered vehicle (U) class licence in South Australia and other states that follow.

Change containment

There are many examples of flexible car platforms for sale today. Volvo sells their small SUV, the XC40 as either full EV or mild-hybrid (fuel powered). Mitsubishi sells their midsize SUV, the Outlander as a plug-in hybrid (electrically powered) or Internal Combustion Engine (fuel powered - ICE) powertrain.?

The “flexible platform” products have created a common notion that building an electric car is as simple as installing an electric motor where a petrol engine goes and Robert is your mother’s brother. The truth is manufacturers that have taken this simple approach have achieved compromised products with expensive production.

So, car makers need to make a truly different type of car, on a truly different type of production line, and that requires a very different approach. The organisational advantages of the pure electric players like Tesla are manifest because teams are working together in ways like never before and it is visible in the product. For example, a legacy automaker will have one team for each of battery (and battery management), motor (and motor management), power systems and cabin HVAC - which results in a typical EV by legacy auto makers having four independent thermal management systems. Tesla has innovated with the ‘Super Bottle ’ an all-in-one heat pump, coolant reserve, thermostat and coolant valve for all of the battery (and battery management), motor (and motor management), power systems and cabin HVAC. The massively reduces production costs and improves packaging efficiency but the people-side to get here is significant.

?Car makers know they need deep organisational change to stay competitive, and this is why we’re seeing the slaughter of sacred cows such as Mercedes-Benz spinning off their commercial division into Daimler, so they can focus on new energy vehicles.

Central stack compute and connected cars

Tesla stock is valued more like a tech company than an industrial player which is a cause of much consternation in both the diamond hand and paper hand community. To the moon. The billionaire investor Kevin O’Leary publicly justified Tesla’s tech-like stock valuation based on their innovation in compute, data, and connectivity. Before I talk about what Tesla is doing (and their competition have committed to pursuing) let me explain the basics and how we got here.

The car shortage caused by a chip shortage highlighted just how many microchips are in a modern car. What happened was this - legacy automakers evolved a complex distributed processing architecture, starting in the 1980’s. This relied on many microchips responsible for various functions placed all around the car communicating with a mix of protocols. These chips were typically off-the-shelf hardware, not necessarily optimised for specific needs and performance demands. The approach was cumbersome, physically bulky, and heavy, complex and created many failure points, it was also not geared for over-the-air updates let alone live streaming operational data.

Microchips in cars grew up like topsy, some new cars from legacy players have over 3,000 of the things. To put that in perspective: if, instead of synthesising crystalline silicone to make chips for cars; one were to synthesise dehydrated, industrially masticated potatoes (42%); high-oleic soybean oil; degerminated, yellow, corn flour; cornstarch; rice flour; maltodextrin; mono- and diglycerides; lactose; and sodium diacetate under comparable pressure, heat, and cleanliness conditions, and form the slurry into stackable, elliptic paraboloid, planar man kibble - as a testament to corporate America's ability to commercialise latent elasticity in our lizard brain’s foraging instinct - the resultant 3,000 chips would require over 50 Pringles tubes for storage.

The alternative is Tesla’s Centralised Processing Unit (CPU) architecture, meaning one powerful central compute stack (safely tucked under the rear seat) handles most of the car's functions - infotainment, autopilot features, battery management, climate control, vehicle diagnostics etc. etc.

Tesla's custom-hardware centralised CPU is tightly integrated with their in-house software, allowing for over-the-air updates and feature enhancements. This approach allows for continuous improvement and innovation without requiring hardware changes.

The result of central stack compute is more capable, flexible, reliable compute on board cars, but it requires a lot more control and capability of the makers to pull it off successfully. The downside is if it's not appropriately hardened, hackers gaining access to one stack, gain access to everything. Conversely, the distributed model has proven even more challenging to harden.

Giga-casting

Pick up the closest Hot Wheels you have handy, flip it over, you are now staring at the future of personal mobility. Yet again, Mattel is over 50 years ahead of the curve, this time in the realm of automotive manufacturing.

What I am talking about is casting the structure underneath your car. Typically, these parts are made of stamping flat sheets of metal and welding them together. Casting has been for small/less complex forms but casting at scale (size) and at scale (volume) is something Tesla mastered and named Giga Casting. They are innovating in more ways than new energy powertrains and touchscreens. They've been leading the charge with their innovative giga-casting technology, first implemented in the Model 3 and Model Y on the road since ~2021.

For Tesla, giga-casting the rear and front third means consolidating over 70 parts and a lot of welding into one single cast which speeds up production and significantly reduces cost.

Tesla's next goal is to cast the entire underbody, potentially streamlining around 400 parts into a single cast. Much like Hot Wheels pulled off were before the USA had landed a man on the moon.

The first mover advantage is wearing off, with reverse-engineering being rife in the auto sector. The reigning world champion of car production, Toyota, has recently announced their intention to adopt this manufacturing approach .

There is risk with this, big ones. If the manufacturers don't figure this out properly, and make a hash of the castings, the results could be ruinous. Manufacturing defects in casting can be difficult to identify and typically cannot be fixed. Cast structural elements usually require a blow torch for removal - any casting defect is a full write-off; any recall will be full car swaps with only scrap value as residual. On the other hand, the risks of not pursuing this are high too, giga-casting goes materially to future cost competitiveness.

So, what does all of this mean for you? The automotive sector is super competitive so cost advantages will be passed on as the industry embraces the change.

However, there are a problem. There have been reports suggesting that these castings may make cars uneconomical to repair after a prang, leading to much higher insurance costs.?

The truth is, these innovative castings have been a part of Tesla models for a few years now, and many owners of Uber drivers I've spoken to have no idea of the significant non-powertrain innovation. Embracing innovation that brings costs of goods sold by significant amounts in a competitive sector is an inevitability.

The end of the dealership sales model

Direct to customer sales, as a concept has been floating around for as long as people have been buying cars. In some places it’s even illegal for the consumer to buy a car without a dealer in the middle of the transaction.?

?Some of you may know, one of the first export markets for Porsche, outside of the German-speaking, world was Australia. The story goes: while on a 1951 European business for the Snow Hydro scheme, Melbourne pump dealer, Norman Hamilton, spotted a very early Porsche model. He visited the factory and cut a deal for the Australasian rights with Ferdinand Porsche himself. The historic partnership ended in 1992 when Porsche AG wrested control of the Australian dealership back from the Hamiltons. Why put an end to a good thing? Because the simple truth is that the dealers’ margin has always been the automakers’ opportunity.

Often times car dealers aren’t family operations anymore. Many dealerships around the world are listed and very profitable. ASX:ASG Auto Sports Group reported EBITDA margin of 8.4% in their 2023 Annual Report.

Beyond the margins, manufacturers require a very different way of interacting with their customers and the old wholesale and markup model isn’t working. This goes to on-going revenue and even many reported problems of dealer staff redirecting potential customers away from electric models. But worse than that, a 2023 survey of car dealers in USA, found 30% wouldn’t sell electric vehicles no matter what.

The franchise dealership model doesn’t seem right for the times and the makers are taking a many different approaches with this (sometimes all at once).

Tesla has a blanket direct sales approach and won’t sell in markets that require a franchise dealer. Porsche tends to avoid franchises where it can.

Mercedes-Benz Australia swapped to a fixed-price direct-to-customer agency sales mode, and had to win a lengthy and brutal legal battle with the dealerships to pull off the change. Honda switched to a similar sales model, but managed to avoid big lawsuits.

Hyundai are trialing Amazon sales of vehicles in the USA and have factory-owned dealers for their new Genesis luxury brand in Australia.

It’s a big and painful change underway, and so far this year 39.7% of battery electric vehicles sold in US were by a? franchise dealers according to NADA.

Higher low voltage

After World War 2, innovations like electric indicators and transistor radios forced the hand of the auto-industry to migrate from 6 Volt to 12 Volt electrics. Since the rise and fall of the Berlin Wall, you would be unsurprised by how much more electrical draw has been installed in cars. Lower Voltage components are cheaper to make, but if you cast your mind back to year 10 physics, the power wastage goes up with the square of current.

Reports are, car makers have been tossing up migrating to 48 Volt since the 1980’s but it was too fiddly - every component in the car that draws power needs to be changed for the new voltage is - every seat heater, windscreen wiper motor, electric seat controller, hazard light button - everything. So why bother? For electric vehicles, wasted power is lost range.

Tesla have committed to 48 Volt architecture starting with the Cybertruck and the rest of their model range will make the change. The improvements are significant. Higher voltage leads to lower current meaning wires and connectors don’t need to be as beefy, resulting in significantly less pricey required per vehicle.

It’s not technically difficult to make the change, it’s just costly and laborious. The business model of most car makers relies on a parts bin , a catalogue of off-the-shelf components they can deploy again and again. Some more innocuous items can lurk in a parts bin unchanged for decades - there’s just no gain in redesigning the little light in a glove box for instance. The kicker is, a swap to 48 Volt necessitates a coordinated redesign of every catalogue item that pull power, this reaches across thousands of tier 1, 2, …, n suppliers. It’s an expensive but soon to be necessary nightmare for legacy automakers.

In the next and final installment, I'll walk through the disruptive mega-trends that aren't led by Tesla.