Table of Contents
Ask ten people where cars are heading and you will hear about flying taxis and egg-shaped pods. Look instead at the production lines, the balance sheets and the city permits actually being issued, and the picture is far more grounded. The future of the automotive industry is being decided by battery chemistry, software margins and how fast a factory can retool, not by concept cars under show-stand lights.
Two numbers set the scene. Global electric car sales passed 17 million in 2024, according to the IEA, which expects EVs to clear half of all new car sales by 2030 under current policies. Battery pack prices fell to roughly $115 per kilowatt-hour that same year, down from about $780 in 2013. Sales climbing while cost falls explains most of what follows.
The car is becoming a software product with wheels attached
A mainstream vehicle now runs on 100 million lines of code or more. Tesla pushed over-the-air updates into the mainstream a decade ago, and the rest of the industry has been catching up awkwardly ever since. Volkswagen set up Cariad to write its own stack and burned through billions before delaying key models. General Motors cut Apple CarPlay from its EVs to keep control of the screen. Rivian replaced dozens of separate electronic control units with a handful of zonal computers, which cut wiring harness weight and let features ship by download instead of by model year.
Follow the money and you see why. Gross margins on mass-market cars often sit in the mid single digits. Connected services, insurance, charging plans and paid feature unlocks can add hundreds of dollars per vehicle per year at far better margins. That is the real contest behind the shift toward software-defined vehicles and the fight for affordability, and it is why so many manufacturers now describe themselves as tech companies, even when the hardware still pays the bills.
Batteries are getting cheaper, and that changes everything
Lithium iron phosphate cells, once dismissed as a cheap stopgap, now power a large share of new EVs, including most of what BYD builds. They give up some range in cold weather but cost less and skip cobalt entirely. Sodium-ion cells are moving from lab papers into small city cars. CATL’s fast-charging packs claim 10 to 80 percent in around fifteen minutes on the right charger.
The awkward part is price. Monthly payments decide purchases, and higher interest rates through 2023 and 2024 hurt EV demand in Europe and the United States far more than any range anxiety did. Several forces are pulling in different directions here:
- Chemistry: cheaper iron-based cells and sodium-ion are pulling entry prices down, especially in China and India.
- Scale: gigafactories are clustering next to assembly plants, cutting freight and inventory costs on packs that once crossed oceans.
- Charging: 350 kW corridors are spreading along motorways, though rural coverage still lags badly.
- Trade policy: the EU applies tariffs of up to about 45 percent on Chinese EVs, and the US has set a 100 percent duty, which pushes Chinese brands toward local plants and joint ventures.
Autonomy arrives street by street, not all at once
Waymo now completes well over 100,000 paid rides a week across Phoenix, San Francisco, Los Angeles and Austin. Those rides are geofenced, mapped in detail and monitored by remote staff. That is the honest shape of the technology in the mid-2020s: superhuman in a known ten-square-mile patch, useless on an unfamiliar rural road in the rain.
Expect robotaxis to spread city by city and robotrucks to take fixed highway corridors between depots first, because predictable routes and depot-to-depot handovers remove the hardest edge cases. Privately owned cars with full hands-off capability on any road remain a 2030s story, whatever the launch events claim.
The factory floor is changing before the showroom does
Walk a modern body shop and you will see the real revolution. Tesla’s 9,000-tonne casting presses turn dozens of welded parts into one piece. Digital twins simulate a line change before a single robot moves, saving weeks of downtime. Supply chain planners track a chip shortage, a fire at a magnet plant and a port strike in the same dashboard, and automakers lean on enterprise resource planning systems that tie procurement, production and finance together to keep a five-figure-part vehicle on schedule.
This is also where the cost war is won. A plant that can switch between petrol and electric models on one line, and between two battery chemistries on the same pack architecture, survives a demand swing that would sink a single-purpose factory.
The skills gap is the bottleneck nobody photographs
Every software-defined vehicle needs embedded engineers, data people and cybersecurity specialists. Every battery plant needs chemists and process technicians. Every dealership needs high-voltage trained mechanics, a role in short supply across Europe and North America. Pay is rising fast for people who can diagnose a thermal runaway risk or debug a vehicle network.
That demand is reshaping education. Two-year technical college programmes in automotive technology now graduate students straight into jobs that pay more than many four-year degrees, and working engineers are topping up with practical AI and machine learning courses to handle driver-assist validation and predictive maintenance.
Follow the capital to see who survives
Building cars has always been expensive. Building electric, connected cars is brutal. Rivian raised billions before turning a gross profit, Lucid has burned through cash chasing range records, and VinFast listed in New York with a valuation that defied gravity for a few weeks. Investors now price these companies on production ramp and cash burn rather than slide decks. If you track high-growth companies heading for public markets, expect the next wave to come from charging networks, battery recycling and automotive software rather than another car brand.
Consolidation is the other half of the story. Smaller brands will license platforms from larger ones, as Ford and Volkswagen have done in Europe, and Japan’s manufacturers are pooling development on software and batteries to avoid duplicating the same research four times.
Signals worth watching over the next twenty-four months
Forget the press releases. These are the measurable markers that will tell you how fast the transition is really moving:
- Average battery pack prices crossing below $100 per kWh, the long-touted threshold for genuine petrol parity.
- The number of cities issuing commercial robotaxi permits, and whether any operator drops the safety driver entirely.
- How much revenue automakers actually disclose from software and services, rather than folding it into vague segment reporting.
- Solid-state battery pilot lines moving from sample cells to production volumes, which would reset range expectations.
- Standardised battery health certificates for used EVs, the single biggest brake on resale values today.
- Whether Chinese manufacturers build European and North American plants rather than shipping into tariffs.
Each of those is checkable, dated and hard to spin. A car bought in 2027 will be judged on how well it receives updates, how its battery degrades after 100,000 kilometres and whether the brand that built it still exists to service it. Those three questions, more than any concept sketch, will define what the industry looks like when the decade ends.
