The Future of Electric Vehicles: Beyond Just Batteries

Jeff Kang

Oct 3, 2025

3 min read

Updated: Aug 21

Thesis: The battery is the least interesting part of the electric vehicle. Charging infrastructure and power electronics decide whether EVs win.

Every conversation about electric vehicles collapses into a conversation about batteries. How many kilometers? How many kilowatt-hours? I used to think this way too, until I watched a relative circle a shopping-center car park twice because two of the four public chargers were marked “out of service” on the app — the car itself charged perfectly once we found a working stall, so the problem clearly had nothing to do with chemistry. That sent me looking at what actually happens between a wall socket and a spinning wheel. What I found convinced me that the industry’s obsession with cells is a distraction. An EV is a power electronics machine that happens to carry a battery, and the reason many drivers still hesitate has almost nothing to do with chemistry. It has to do with whether they can add 300 kilometers of range in fifteen minutes on a highway they have never driven before. That is an infrastructure problem which belongs to electrical engineers.

The core technology

Inside an EV, direct current from the pack must become an alternating current for the motor, and it must do so thousands of times per second with almost no loss. That job belongs to the traction inverter, which switches high-power transistors on and off under software control. The shift from silicon to silicon carbide has been quietly transformative: wide-bandgap devices tolerate higher voltages and switch faster with lower losses, which is worth several percent of total range without adding a single cell. Around this sit the DC-DC converter, the onboard charger, and the battery management system, which monitors cell voltages and temperatures, balances the pack, and estimates state of charge. Efficiency at every one of these stages compounds.

*Figure 1. SiC-based traction inverter power module (pin side / heat-sink side)*

Recent developments and real problems

The most meaningful recent change is not a new chemistry but a new voltage. Moving vehicle architectures from 400 volts to 800 volts halves the current for the same power, which cuts resistive losses and allows lighter cables. This is what makes 350-kilowatt charging physically reasonable. Meanwhile, charging standards have painfully strengthened after years of fragmentation, and vehicle-to-grid capability is beginning to appear. Yet the honest problem remains reliability. Chargers that are broken or throttled do more damage to EV adoption than any range figure. A driver who has been stranded once will describe the technology as immature regardless of how good the pack is. Charging is also a siting and permitting problem as much as an electrical one, since a high-power site needs a grid connection, and often local storage, that utilities can take years to approve.

*Figure 2. A Porsche Taycan fast-charging*

Outlook and solutions

Solid-state batteries will arrive eventually, and they will be excellent. But they will not fix a charging network that nobody maintains. The higher-leverage solutions are unattractive: mandatory uptime reporting for public chargers, buffered charging stations that use local storage so a site does not need a massive grid connection, and continued adoption of wide-bandgap semiconductors in both vehicles and chargers. Bidirectional charging deserves particular attention, because a parked fleet is a distributed storage asset that could stabilize a grid absorbing more solar every year.

Standardizing the communication protocols between vehicle, charger, and grid operator is the missing piece, and it is a standards problem rather than a scientific one.

*Figure 3. Vehicle-to-Grid (V2G) bidirectional charging concept*

Conclusion

My view is that the winner of the EV transition will not be whoever announces the best cell. It will be whoever builds the most reliable charging experience on top of the most efficient power electronics. I recognize the counterargument: a genuine leap in energy density would shrink the charging problem simply by requiring less of it, and the researchers chasing that leap are not wrong to try. But every serious density roadmap I read still assumes today’s charging network, and that network is the part actively failing drivers right now. If I study electrical engineering, this is where I would want to work, not because inverters and connectors are exciting to talk about, but because they are the actual bottleneck.