Electric Vehicle Engineer Interview Questions

“I got hooked in my second year when I joined the college electric racing team. I started on wiring harnesses, but I kept asking why the pack limited our power on hot days, and that pulled me into the battery side. By final year I was running the team's pack testing, logging cell temperatures and tuning the current limits with our BMS supplier. During my internship I worked on thermal testing of a battery module, which confirmed it for me. So the part I want to build a career in is battery systems, especially thermal management and the BMS logic around it, because that's where range, charging speed and safety all meet.”

“In my final-year project we built a small battery pack for an electric scooter, and the problem was that it cut out on hills even with charge left. My part was the analysis. I logged cell voltages and current on a few hill runs and saw that one parallel group sagged much lower than the others under load, so the BMS hit its undervoltage limit early. We traced it to a poorly welded tab with high resistance. I reworked the welding process with the lab technician and added a resistance check for every group before assembly. After the fix, the voltage spread under load dropped to a small fraction of what it had been and the scooter climbed our test hill without cutting out. Next time I'd put that resistance check in from day one.”

“At the heart is the high-voltage battery pack, with its BMS and contactors that connect it to the rest of the car. From there a distribution unit feeds the traction inverter, which turns DC into three-phase AC for the motor, and the motor drives the wheels through a single-speed reduction gear. The same high-voltage bus feeds the onboard charger, the electric AC compressor and the cabin heater. A DC-DC converter steps the high voltage down to keep the 12 volt battery and all the control units running. A vehicle control unit coordinates everything. When I drive, power flows from the pack through the inverter to the motor. When I lift off or brake, the motor works as a generator, the inverter sends that current back into the pack, and the BMS decides how much regen it can accept, which is less when the pack is cold or nearly full.”

“A BMS measures every cell group's voltage, the pack current and temperatures around the pack. It uses that to protect the cells: if any cell goes too high, too low, too hot or the current is too big, it first cuts the allowed power and, if needed, opens the contactors. It balances cells, tells the vehicle and the charger how much power they may use right now, and estimates state of charge and state of health. State of charge can't be measured directly. Counting current in and out works well short term, but any small sensor offset builds up. Looking up charge from a rested cell's open-circuit voltage corrects that, but you rarely get a fully rested cell while driving. So a good BMS combines both, often with a cell model and a Kalman-type filter that trusts each source when it's most reliable.”

“Series connections add voltage, so 96 cells in series at 3.6 volts gives 345.6 volts nominal. Parallel connections add capacity, so three cells in parallel at 50 amp-hours gives 150 amp-hours. Energy is voltage times capacity, so 345.6 times 150 is 51,840 watt-hours, about 51.8 kilowatt-hours nominal. That's 288 cells in total. I'd add that the usable energy is lower, because the BMS keeps a buffer at the top and bottom, and the voltage swings between roughly the cells' minimum and maximum across the charge range, so the bus the inverter sees might run from under 300 volts up to around 400.”

“With AC charging, the car gets AC from the grid and its own onboard charger converts it to DC for the pack, so the power is limited by that onboard charger, typically a few kilowatts up to around 22 kilowatts on three-phase. With DC charging, the big converter sits in the station, and DC goes straight into the pack through the car's DC contactors. The BMS tells the charger what voltage and current it wants at every moment. Early on the charger can push a high, steady current. As the cells approach their maximum voltage, the BMS has to reduce the current so no cell overshoots its limit, and that's also where lithium plating becomes a risk, especially if the pack is cold. That's why the last part of a fast charge takes much longer, and why fast-charge times are usually quoted only up to about eighty percent.”

“An EV's drivetrain is very efficient, so the big losses are out on the road. Aerodynamic drag force grows with the square of speed, so going from city speeds to highway speeds makes drag the biggest energy user, and the energy per kilometre climbs quickly. Rolling resistance stays roughly the same per kilometre. Cold hurts in a few ways. Heating the cabin has to come from the battery, because there's no engine waste heat, and in slow winter traffic that heater runs for a long time per kilometre. A cold battery also has higher resistance, so less of its energy is usable, and the BMS limits regen until it warms up. That's why heat pumps, battery preconditioning while plugged in, and good aerodynamics make such a difference. An engine car loses efficiency too, but it has free heat and its losses are already large, so drivers don't notice.”

“In my internship I helped swap a module on a test pack at around 400 volts. I'd done the company's HV awareness training, and I worked under a qualified technician. We isolated the pack by pulling the service disconnect and locking it out with our own tags, then waited the specified time for the capacitors to discharge. The technician checked our meter on a known source, measured for zero volts between the terminals and to the chassis, and checked the meter again afterwards. We used insulated gloves that had been checked that day and insulated tools, and we covered any exposed terminals we weren't working on. The part I took away is that the checks happen every time. Even when the pack is known to be off, you prove it's dead before you touch it.”

“From what I've read, you build electric delivery vans for city fleets, so the vehicles do lots of short trips, stop-start driving and depot charging every night. That makes battery life, cost per kilometre and uptime matter more than top speed. That's exactly the kind of problem I like. In my last role I worked on engine calibration, and I enjoyed tuning for efficiency, but with an engine most of the big decisions are already made. With an EV the battery, motor and charging strategy are still being shaped, and a good engineering choice shows up directly in how long a fleet vehicle lasts and what it costs to run. I'd like my work to land there.”

“LFP, lithium iron phosphate, stores less energy per kilo than NMC, so for the same range the pack is heavier and bigger. In return it's more thermally stable, it usually lasts more cycles, it avoids nickel and cobalt so it's cheaper, and it's comfortable being charged to full every day. The downsides are poorer cold-weather performance and a very flat voltage curve, which makes state of charge harder to estimate. NMC, nickel manganese cobalt, packs more energy into the same space, which suits long-range or performance cars, but it has less thermal margin and is usually kept below full for daily use to protect its life. So for a city car, a fleet van, a bus or a cost-sensitive two-wheeler I'd lean towards LFP. For a long-range car where weight and space are tight, NMC.”

“A permanent-magnet motor gets its rotor field from magnets, so it doesn't waste energy creating that field. That gives high efficiency, especially at the part loads where cars spend most of their time, and lots of torque for its size, which is why it dominates. The costs are the rare-earth magnets, which bring price and supply risk, magnets that can lose strength if they get too hot, and a back-EMF that rises with speed, so at high speed the inverter has to weaken the field. An induction motor has no magnets, it's rugged and cheaper in materials, and when it isn't powered it spins freely with almost no drag. That makes it a sensible choice for a second axle that only helps under hard acceleration. Its downside is lower efficiency and rotor heat that's hard to remove.”

“For AC, the common plugs are Type 1 in parts of North America and Japan and Type 2 across Europe and many other markets, with GB/T in China. For DC, CCS1 and CCS2 add two DC pins below those AC plugs, CHAdeMO came from Japan, GB/T has its own DC connector, and North America has standardised the SAE J3400 connector. Some markets also have national standards for smaller, lower-voltage vehicles, Bharat DC-001 being one example. Underneath, AC charging is simple: the charger puts a PWM signal on the control pilot line and its duty cycle tells the car the maximum current it may draw. The car's onboard charger stays under that. DC needs a real conversation. On CCS, the car and charger exchange digital messages over powerline communication on the pilot line, while CHAdeMO and GB/T use CAN. The BMS keeps requesting a voltage and current, and the charger follows.”

“First I'd stop the charge, because the heat source that matters most is the current going in. I'd keep the cooling loop running, since that's what pulls heat out, and I'd follow the lab's emergency procedure: alert the test lead, keep people back, and have the right suppression and venting ready in case it's heading towards runaway. I wouldn't open or move the pack until it's stable and cooling. Once it's safe, I'd save every log before anything is reset. Then I'd work out whether it's real heat or a bad reading, by comparing neighbouring sensors and the cell voltages in that module. If it's real, the usual suspects are a cell or joint with high resistance, a blocked or air-locked cooling channel under that module, or a poorly seated thermal pad. Nobody retests that pack until we know the cause.”

“At my last company, a prototype pack kept throwing an overtemperature fault during fast-charge tests. Everyone assumed the cooling plate was undersized, and a redesign was being discussed. Before that, I lined up the logs and noticed the fault always came from the same sensor, and the neighbouring sensors never agreed with it. The coolant temperatures in and out also looked normal for the heat we were putting in. I asked the technicians to fit extra thermocouples near that sensor, and they read several degrees lower. When we opened it up, the sensor's thermal pad had been fitted crooked, so it was reading heat from a busbar joint instead of the cells. We fixed the fitting and added a check to the assembly instruction. It saved a cooling redesign that wasn't needed.”

“I try to go to sources that have to be precise, like the standards themselves, technical papers, supplier application notes and teardown reports, rather than press releases. When something new appears, say a new cell chemistry or a faster charging claim, I ask two questions. Is it proven in production at scale, or only in a lab? And if it's real, what would it change for the vehicles we build? If the answer is not much for the next few years, I'll keep an eye on it but won't go deep. If it touches our product, like a new version of a charging standard we must support, I'll learn it properly, and I'll usually write a short summary for the team so we don't all read the same hundred pages. Honestly, most of my learning comes from test data on our own packs.”

“I'd take the saving seriously, because cell cost drives pack cost, but I'd explain that two datasheets matching doesn't mean two cells behave the same. Cells can differ in resistance, how they age, how they respond to fast charging and cold, and how they behave in abuse tests. Every one of those feeds the BMS models, the thermal design and the safety case. So I'd propose a plan: get samples, run our own characterisation and start ageing tests right away, repeat the key abuse and propagation tests, audit the supplier's manufacturing and quality data, and update the BMS parameters. Then I'd give purchasing an honest timeline. If it can't be done before launch, the practical answer is usually to launch on the current cell and bring the new one in later as a validated change, rather than rush it.”

“I'd be clear that I can't just change the number, because the current limits come from cell testing. They're set to avoid lithium plating, overheating and fast ageing across temperatures and states of charge, and the safety case and warranty assume those limits. Raising them without testing could mean packs that fade early or, worst case, plating that raises the risk of internal shorts. But I wouldn't stop at no. I'd look at where the charge curve is being held back. Often there's room in specific windows, like the middle of the charge when the pack is warm, and better preconditioning before arriving at the charger can gain a lot. I'd propose a quick test plan on real cells to find that headroom safely, with a date. If we still can't beat the rival safely, I'd say so plainly and put it in writing.”

“The inverter has three half-bridges, one per motor phase, six switches in total, with a DC-link capacitor across the battery side. By switching each leg thousands of times a second with PWM, it builds average voltages that look like a three-phase sine wave, and it can change their frequency and amplitude freely. On top of that sits field-oriented control. It reads the phase currents and the rotor angle from a resolver or encoder, transforms the currents into a frame that rotates with the rotor, and controls two parts separately, one that makes torque and one that sets the field. Silicon carbide MOSFETs switch faster and with lower losses than silicon IGBTs, especially at light load and at 800 volts, so the inverter wastes less energy, needs less cooling and adds range. The trade-offs are higher cost and faster voltage edges that stress motor insulation and create more EMI.”

“First I'd narrow it down: which charger models and software versions fail, whether every car of ours fails or only some software versions, and at which stage the session dies. A DC session goes through clear steps: plug detection, the digital handshake, insulation check, pre-charge, then current flow. Our charging logs usually show where it stopped. Then I'd capture a full trace of the conversation at a failing charger, and ideally get the charger's logs from its operator. Often it's a timing difference, like the charger expecting a response faster than we send it, or both sides reading an optional part of the standard differently, or the insulation test tripping because of something on our side. Once I know which message or step fails, I'd check it against the standard. If we're wrong, we fix our software. If they are, I'd share the trace with the charger maker, and consider a workaround meanwhile.”

“I'd start from what the cells need. They're happiest near room temperature. Too hot and they age faster and move closer to thermal runaway. Too cold and their resistance climbs, so power drops, and charging fast risks lithium plating. Then I'd look at the duty cycle: how hard the car is driven, how often it fast-charges, and the hottest and coldest markets it's sold in. That sets the heat I need to remove and add. A small, low-power vehicle might manage with air cooling, but most cars use liquid cooling plates under or between the cells, with a chiller linked to the AC system for hot days and a heater or heat pump for cold ones. A key target I'd set is a small temperature spread between cells, because uneven temperature means uneven ageing. I'd model it early and then confirm on the bench with lots of thermocouples.”

“Thermal runaway starts when a cell heats faster than it can shed heat, maybe from an internal short, overcharging, an outside fire or crash damage. As it heats, the protective layer on the anode breaks down, the separator melts and the electrolyte decomposes, and each step releases more heat. The cell vents hot, flammable gas and can reach temperatures that ignite its neighbours. To stop that spreading, I'd work in layers. Prevent it with good cell quality and BMS limits. Slow the spread with heat barriers like mica or aerogel between cells or modules, and spacing where the design allows. Give the gas a planned vent path out of the pack, away from the cabin. Use fusing so a shorted cell doesn't dump its neighbours' energy into it. And detect it early from temperature rise, voltage drops or gas and pressure sensors, so the car can warn occupants and give them time to get out, which several regulations now require.”

“I'd start with consumption on the real route, not a lab cycle. Say we measure about 250 watt-hours per kilometre with a typical load. Winter heating and a cold battery might make that about a third worse, so I'd plan on roughly 330, which is about 50 kilowatt-hours usable for 150 km. Then two margins. The pack will fade over its life, so if we promise the range until it's at about eighty percent of its new capacity, it needs around 62 kilowatt-hours usable when new. And the BMS keeps a buffer at the top and bottom, so the gross pack might be close to 70. After that I'd check the knock-on effects. A bigger pack is heavier, which raises consumption and cuts payload, so I'd iterate once. And I'd ask whether the vans could top up at the depot at lunchtime, because that could let us use a smaller, cheaper pack.”

“On a two-wheeler programme we were short of range, and the obvious fix was to add more cells, but that added weight and pushed the cost over target. I built a simple model of consumption against weight and compared three options: more cells, a better-matched motor and gear ratio, or lower-rolling-resistance tyres plus a tweak to the regen map. The model showed the motor and gearing change gave most of the range we needed without the weight. I took the numbers to the programme manager and the chassis team, because the tyre option affected grip, and we agreed to test the gearing change first. On the road it got us within a few kilometres of the target, and the regen tweak closed the gap. What I held firm on was the cell temperature limits. Nobody suggested it outright, but running the cells hotter would have been a hidden way to buy range.”

“Two months before a design freeze, our winter range tests came in clearly below the target we'd promised. I rechecked the setup first, because I didn't want to raise an alarm over a test error, and the repeat run agreed. I told my manager the same day and wrote a one-page summary: the measured numbers, where the energy was going, mostly cabin heating and a cold pack, and three options with rough effort for each. They were better preconditioning logic, a heat pump the programme had dropped earlier, or updating the published range. It wasn't a comfortable meeting, but because we had the data and options, the discussion was about choosing, not blaming. We went with the preconditioning change plus revised winter guidance for customers. I learned that bad news with options is received very differently from bad news alone.”

“I'd start with data before assuming anything. From connected vehicles or service tools, I'd look at energy used per kilometre against outside temperature, how much went to cabin and battery heating, trip lengths, and each pack's state of health. Most winter range loss is normal: short trips in the cold spend a lot of energy warming the cabin and battery, and a cold pack gives up less energy. If these cars match what our own cold tests predicted, the issue is expectations, not hardware. But I'd watch for outliers. A car losing far more than others in the same climate might have a heater stuck on, a heat pump fault or a real capacity drop. Those go to service and root cause. For everyone else, I'd work with the customer team on clear guidance about preconditioning while plugged in, and see whether a software update could make heating smarter.”

“The inverter and other loads have large DC-link capacitors. If you closed the main contactors straight onto them while they're empty, you'd get a huge inrush current that can weld the contactors and stress the capacitors. So the BMS first closes the negative contactor and a pre-charge path through a resistor, waits until the bus voltage is close to the pack voltage, then closes the main contactor and opens the pre-charge. It also checks the timing: too fast or too slow points to a fault. The interlock loop is a low-voltage circuit threaded through every HV connector and service cover. If one is unplugged or opened, the loop breaks and the system opens the contactors. Isolation monitoring exists because the HV system floats, with no connection to the chassis. One fault to the chassis isn't immediately dangerous, but a second one could be, so the system measures that resistance constantly and warns or shuts down when it drops too low.”

“Functional safety is about making sure a fault in an electrical or electronic system can't lead to unreasonable risk, and in cars the main standard is ISO 26262. It starts with a hazard analysis. For each hazardous event you rate how severe the harm would be, how often the driving situation happens, and how well a driver could control it. Those three ratings give an ASIL from A to D, or QM if normal quality processes are enough. For a battery, one safety goal would be to prevent cell overcharge that could lead to thermal runaway. That usually lands at one of the higher ASILs, so you need safety mechanisms like a second, independent way to measure cell voltages, a hardware path that can open the contactors even if the main software fails, and a defined safe state. Then every requirement is traced down to hardware and software and verified.”

“I'd plan it in layers. At cell level, capacity and resistance across temperature and charge, plus ageing tests that run for months, so they start first. At pack level, electrical performance, thermal tests including repeated fast charges, vibration and mechanical shock, water and dust ingress, thermal cycling, and abuse tests like overcharge, external short and propagation from a forced runaway. The BMS gets tested on a hardware-in-the-loop rig with a cell emulator, so I can inject faults that would be too dangerous on a real pack. Then EMC, and vehicle tests in hot and cold chambers and on the road. Before any of that, I'd map the rules for each target market, because they differ: UN 38.3 for shipping, UN ECE R100 in many countries, and national rules such as GB 38031 or AIS-156. I'd link every test back to a requirement, so gaps show up early.”

“On one prototype, the car would sometimes refuse to close its contactors on a cold morning. The BMS supplier said it was a vehicle wiring issue, our harness team said it was the BMS. I asked for one shared log from a failing start, with the BMS and the vehicle controller recorded together. It showed the pre-charge was timing out. The BMS waited for the inverter to report its bus voltage, and on cold mornings the inverter's controller woke up later, so that message arrived after the BMS had given up. So it was really a requirement gap between us and the supplier, not anyone's mistake. I wrote up the finding with the traces, got the supplier's firmware lead, our harness engineer and the vehicle software owner on one call, and we agreed a revised timeout plus a wake-up order that starts the inverter controller first. I tracked it to closure and we confirmed it in the cold chamber.”

“I'd want them to raise it straight away, with me directly or through the formal issue process if they're more comfortable with that. Either is fine. The worst outcome is someone sitting on a concern because the design has a senior name on it. When they come to me, I'd thank them first and mean it, then sit down and go through their reasoning and the data together. If they're right, I'd own it openly, reopen the design and make sure they get credit in the review. If they're not, I'd explain why with the same care, walk them through the analysis, and still make it clear I'm glad they asked. In high-voltage work, one ignored concern can hurt someone, so I'd rather hear ten false alarms than miss one real one.”