When you can see the batteryused EVs are easier to buy and sell
Buyers worry about a worn-out battery. Sellers worry that a good car can't show it's good. CELLPROOF comes to the car and puts the battery's condition on a certificate tied to the VIN, which anyone can verify online.
Used cars are judged by mileageEVs should be judged by the battery
A listing's mileage tells you how far the car has gone. It can't tell you how the battery has held up.
2021 Tesla Model 3 RWD
- Mileage
- 48,200 km
- Battery
- ?
Same 50,000 km, same battery?
Not necessarily. Two EVs with the same 50,000 km can have very different batteries:
Charging habits
A car that fast-charges often, or is often charged to full, ages its battery differently from one charged slowly.
Taiwan's heat
Long hours parked in the sun and heat wear a battery too.
Time itself
A battery ages slowly even when the car sits. Low mileage doesn't mean a young battery.
Geotab, a fleet telematics company, followed more than 20,000 EVs: cars that rarely use DC fast charging lost about 1.5% of their battery a year, while cars that often use 100 kW+ fast charging lost up to 3% a year.
What about the battery health the car shows?
The car keeps its own record of the battery, and that's a good starting point. Before you read that number, there are three things to know:
Sample car
In the sample car, the car records 98% and DEEP measured 94%. One is an estimate the car builds up over time; the other is the energy measured that night. Side by side, you see more. This car's gap is within the normal range.
01
Is the range on the dashboard the battery's capacity?
No. The range on the dashboard is projected from how you've been driving lately. Drive faster, run the air conditioning, or let the weather cool, and it moves. It tells you how far you can go, not how much energy the battery still holds. So two identical cars 30 km apart at full charge don't necessarily have batteries that far apart.
02
Can you compare 92% across brands?
Not directly. Each brand works out health its own way: some look only at capacity, some fold in internal resistance, and they update at different times. 92% on one brand and 92% on another are not the same thing. Comparing a car with itself over time is what makes sense.
03
Can a screenshot tell you which car it is?
No. A screenshot can't show which day it's from or whether it's even this car. That doesn't mean the seller is hiding anything; a screenshot simply can't be verified. The certificate ties the number to the VIN, and one scan verifies it.
What does CELLPROOFactually measure?
CELLPROOF has two tests, answering two questions: how is the battery doing right now, and how much energy does it actually hold now? It isn't a cheap version and a proper one.
SWIFT
5 minutes per car, a full battery check-up
Reads the complete data from the car's battery system: the health the car records itself, whether each cell group is balanced, temperature spread, fast-charge share, lifetime charging, the low-voltage battery and the factory warranty left. Fast and complete, all on one certificate.
The first step when viewing or listing a car: 5 minutes to see the battery clearly.
DEEP
One night, the capacity measured by hand
Everything in SWIFT, plus a real measurement over one full charge and a high-voltage insulation check, to find out how much energy this battery actually holds now.
Before you spend big, this is the number that lets you buy with confidence.
| SWIFT | DEEP | |
|---|---|---|
| The question | How is the battery doing right now? | How much energy does it actually hold now? |
| How | Reads the complete battery system data | Everything in SWIFT, plus AC charging from below 20% to 100%, recorded throughoutAbove 20% still works, with lower confidence; above 50% we can't run a DEEP test. |
| Checks | Cell balance, temperature spread, sensors, system status, battery alerts during the test, low-voltage battery | Everything in SWIFT, plus high-voltage insulation and measured charge energy |
| Battery health | The car's system value | Measured by CELLPROOF, beside the car's own figure |
| Certificate and online check | Yes | Yes |
| Time | 5 minutes per car | One night (connected in the evening, collected in the morning) |
To see the battery's condition quickly, choose SWIFT. To know the actual capacity, choose DEEP. Both come with a certificate you can verify online.
See plans and prices →How can CELLPROOF help me?
Pick the one that fits you, then open each item for the details.
Customers ask about the battery more and more. Have the answer ready: buy in with confidence, and list cars that are easier to sell.
Check the battery before you buy in
When you buy in a used EV, the battery is the biggest unknown. Run a SWIFT first, 5 minutes per car, and you know where you stand before you price it, with no fear of pricing it wrong.
Put the certificate on the listing
Every test comes with a preview certificate image you can post straight onto listing sites such as 8891. Buyers see the health, grade and estimated range at a glance, and can verify it online with the certificate number.
Stop answering the battery question car by car
The battery is the question buyers ask most. With a certificate attached, you don't have to explain it car by car or send screenshot after screenshot.
Visitors scan the test card themselves
The test card sits inside the windscreen. A customer viewing the car scans the QR code and opens the full report: the cell map, charge curve and warranty left are all there.
A third-party number puts customers at ease
The number is measured by a third party that takes no part in the sale and is tied to the VIN, so customers feel safer. We take no commission, and the result doesn't change with who ordered it.
Lots of cars? We test a batch on site
We come to your lot and test a batch in one go, SWIFT or DEEP. Volume pricing from 5 cars, every result in one spreadsheet, a monthly summary report, and dealer test cards.
Why should you trustour numbers?
Is it safe to plug your equipment into my car?
We use professional European automotive data logging equipment. Its maker's equipment is used by more than 10,000 companies worldwide, including automaker engineers and research institutions.
Receives only
Sends nothing, not even an acknowledgement
In our configuration the equipment only receives and never transmits, not even an acknowledgement. It cannot send commands to the car or change any setting. It simply listens.
The equipment has passed:
- UN ECE Regulation 10 automotive EMC testing, performed by TÜV SÜD, covering emissions, immunity and supply-line transients
- ISO 7637-2:2011 automotive voltage transient testing, performed by TÜV SÜD, covering transient emissions and immunity on supply lines
- EMC testing in accordance with CE (EU), FCC (US) and IC (Canada) standards
- An EU Declaration of Conformity covering the EMC and RoHS directives
- The EN 62368-1 electronic equipment safety standard
Are you on the buyer's side or the seller's?
We're on the side of the numbers.
Whether the car sells, and for how much, we're paid the same. We don't list cars, broker deals, buy or sell cars, or take commission.
Independence Pledge
We promise to stay independent, always.
- We take no part in any vehicle sale.
- We take no referral fees, kickbacks or sales commission from dealers.
- A result does not change because the client is a buyer, a seller or a dealer.
- The numbers reflect the measurement only. We give no buying or selling advice.
A certificate only means something when buyers and dealers can both trust it.
What we can and can't measure
What we can measure
- Measured charge energy (DEEP)
- The gap between the car's own figure and ours (DEEP)
- Cell voltage and temperature spread
- High-voltage insulation (DEEP)
- Battery alerts the car broadcasts during the test
- Low-voltage battery condition
- Lifetime charge and fast-charge share
What we can't
- How long the battery will last
- Energy delivered at high current on the road (we measure under AC charging; driving gets a little less)
- Mechanical damage or external factors
- Accuracy and error range (comparison data still building; not published yet)
- Whether the manufacturer will honour a warranty claim
A certificate speaks only for the day of the test; it isn't a promise about lifespan or range. Whether a warranty claim is accepted is the manufacturer's call.
When will you publish accuracy?
When there's enough comparison data. We're building it three ways:
- Testing the same car again after a while, to check the method is stable.
- Testing both near-new cars and high-mileage commercial cars, to check both ends make sense.
- With the owner's permission, running the manufacturer's own battery test on the same cycle, recording both results, and publishing the comparison once there are enough.
Until there's data, we won't quote an error range.
Which cars can you test now?
For now, selected Tesla models only; other brands are in preparation. The full list is on supported models. Air temperature, starting charge and the charger all affect a measurement. We aim for similar conditions every time, and if conditions are clearly poor, we'll suggest another day.
How does DEEP measure?
What DEEP does is simple, really: it stays with your car through a full night of charging and works out, from the most basic voltage and current, how much energy this battery holds now. The hard part is that every step has its catches.
- 01
Connect, and only listen
We read the raw signals from the car's diagnostic port and send no commands.
- 02
Start below 20%
The charge comes down below 20% first. Above 20% still works, with lower confidence; above 50% we can't run a DEEP test. We connect on site in the evening and collect in the morning.
- 03
AC charge to 100%
A full night of charging, every stretch of voltage and current recorded. The moment it's full is the most reliable reference point of the whole measurement.
- 04
Work it out
We correct the sensor offset, work out the measured charge energy, and print it beside the car's own figure.
One night, hour by hour
Sample test record
21:02
Connected
Equipment on, the battery rests first
23:33
Charging starts
AC charging, recorded throughout
06:28
Full
Cell voltage reaches its limit, charging stops
08:41
Analysis done
Rested again after full, result worked out
Just reading the signals is a hurdle
The car's computers talk to each other over the CAN bus, thousands of messages a second. None of it is publicly documented, and the format changes with the model, the year, even the software version. Misread one signal and everything calculated after it is wasted.
The sensors tell small lies
Even with no current flowing, the current sensor often reads a little. We use rest periods to find that offset and take it out.
Read more
On its own it looks like nothing. Added up over a few hours, it can become more than a kilowatt-hour of error, and nothing in the result gives it away. That's why we rest before and after. That's the difference between adding numbers up and actually measuring.
How a tiny offset becomes a large error
IllustrationSo why not fit our own sensor? An external current sensor means working on high-voltage wiring, and it brings calibration problems of its own. So we deliberately use the sensors already in the car, recalculate from the raw signals and correct for their own offset.
Why does it have to go all the way to 100%?
To work out capacity you need to know how much the charge changed. The percentages in between are the car's own calculation; the moment the battery is full, its cell voltage reaches a physical limit, the most reliable reference point there is. That's also why DEEP takes a night.
Read more
Measure only from 20% to 80% and both ends are calculated values, so the capacity you get is uncertain too. Full isn't an estimate; it's physics.
Why 100% is the reference point
IllustrationDoes the same calculation work for every battery?
No. Nickel-based (NCA, NMC) and lithium iron phosphate (LFP) batteries behave completely differently in voltage. We handle each chemistry and pack specification separately.
Read more
An LFP battery's voltage is almost flat through the middle. Full, at 100%, is just as reliable, but at the low end NCA and NMC batteries still let voltage cross-check the charge level, and LFP is too flat for that. So LFP results carry more uncertainty, and the certificate always states the chemistry.
Why LFP is harder to measure
IllustrationDo heat, cold and cell differences change the result?
Yes, so we record temperature the whole way and take it into account.
Read more
A battery gives up different amounts of energy at different temperatures. Once it's full, the BMS starts balancing the cell groups, and that process says something about the inside of the pack. The true gap between cells only shows once the battery has rested and its voltage has settled.
Taiwan's seasons swing less than Europe's, but a summer garage and a winter night still differ. Once there's enough data, we'll adjust results to one reference temperature.
We keep the raw record of every test. Every number on a certificate traces back to data we actually measured.
The numbers on the certificatehow do I read them?
Tap an outlined block on the certificate to see what it tells you.
Sample1 / 6
1Number and test
A dark card is DEEP, a white outlined one is SWIFT. Enter the certificate number on the verify page to confirm it's real and still valid: a certificate is valid for six months from the test date.
Will it alwaystake a whole night?
Right now the only way to know a battery's real capacity is an overnight DEEP test. We want to change that.
Every DEEP test becomes part of a Taiwanese battery dataset in which every entry is a real measurement. Once there's enough of it, we'll be able to estimate a DEEP result from a SWIFT test. That estimate will be built entirely on Taiwan's cars, Taiwan's weather and the way people here drive.
Before anything goes into the dataset, names, contact details and the full VIN are removed. If you'd rather not take part, just tell us when you book; it doesn't affect your test or certificate.
Is this done elsewhere?
Independent battery health testing has been around in Europe and the US for years. AVILOO in Austria and Voltest in the US have both made it part of buying and selling used cars.
We learned from that: keep the quick check and the full measurement separate, define health by measured charge energy, and show the car's own figure apart from the independent one.
We have no relationship with either company. We're a Taiwanese team, focused on Taiwan's cars, Taiwan's climate and Taiwan's used-car market.
Buying, selling or listing a car?See the battery clearly first
The tester comes to you: SWIFT takes 5 minutes per car, DEEP one night.