_{March 23, 2026 // Blog // EN}

Battery Data: Are You Seeing the Full Picture? [The Battery Cycle #7]

In 2021, Claudius Jehle, CEO of volytica diagnostics GmbH, and Sustainable Bus launched a series of articles around the “The Battery Cycle”. They shed light on the complexities of Li-Ion batteries and provided valuable insights for anyone involved in electric mobility. In 2025 and 2026, the articles were updated with real-world data and lessons learned to understand how battery chemistry affects real-world operations.
For questions and contributions please reach out to us via contact@volytica.com.

Setting the Stage: Why “SOH” Depends on Battery Data and Definitions

“Data is the new oil”, people keep saying. But like with loads of crude, thick, stinky and gooey oil, how to distil value out of it is the real question. This article shall put a general light on the question of data, ownership, usage rights, and more specifically, the trustability of data and what can be extracted from it.

SOH is a calculated health indicator, not a directly measurable physical quantity. — SOH depends on input data quality, resolution, and the evaluation method applied.

When we speak about “data”, we generally mean every information that comes from, or is associated with, an asset like an electric bus. This entails meta information like data sheets and technical parameters that are (almost) not changing over time, like type-plate capacity, used battery cell type, the manufacturer of the doors, etc.

But the real amounts come from time series data, that is, sensor readings collected during operation of the vehicles. There are more straightforward signals like the passenger compartment temperature, the number of passengers, or whether the doors are open or not (“yes/no”).

But if we keep focusing on the battery alone one realizes how much battery data these systems actually handle every minute: Every cell in a battery pack has one voltage sensor that is read at least 10 times per second. Similar with current and temperature sensors. So a battery produces thousands of sensor readings per minute, and gigabytes per month. Thanks to many initiatives like the VDV 238 quasi-standard, today more often than not this raw oil is accessible by attaching the right hardware to the communications interfaces of buses, like the FMS. Often, this information is also transmitted to a Cloud.

Electric heavy duty vehicles generating operational battery data — Electric heavy-duty vehicles generate large volumes of operational data during daily use, forming the basis for independent battery analysis and performance assessment.

And an increasing number of professional public transport operators (PTOs) are starting to unlock this treasure trove of information by employing tools to automatically translate crude oil, erm, data into key performance indicators like “SOH” (make sure you read our article on SOH!), safety indicators or other crucial information to extract the maximum from the batteries and use these black boxes to their true potential.

Data Ownership

But in doing so, the big question arises frequently: Who is actually owning the raw data? And if someone else than the OEM extracts KPIs, are they better or more trustable or “legally usable”? What if my own SOH assessment differs from the OEM’s? What if I can’t get hold of the necessary raw data to make such independent own assessment in the first place?

Let’s look at ownership. The sad news is—there is no such things as “data ownership” (in the EU). Data is legally not “owned”, in the sense of: Only one party can do something with it. Rather, on data there are usage rights.

So the question should rather be: Who has the right to use which kind of data, and subsequently, who is allowed to access this information in the first place? Until recently, this was all but clear, but the EU Data Act from 2025 has certainly set new standards here, effectively giving the owners a prominent (a) right to access, and (b) right to use and also share vehicle data with 3rd parties (no such legal framework exists in the US, to be fair)!

Case closed, everyone can access everything, by law? As always, there is plenty of room for debate how to interpret, “live” and enforce this. Particularly as not everyone would want to go into a legal escalation over this important, but – at least as many OEMs argue – not vital question.

So irrespective of the abstract legal terms and “Acts”: In the business and professional context we are all operating, access and usage rights to data are more a negotiation and buying power topic, rather than something to appeal towards the OEMs e.g. after a purchase order is placed. In other words: There are ample examples of smart transport and logistics operators that successfully strong-armed manufacturers into granting sufficient access to the relevant data, without appealing or relying on law alone.

Who is Right and Trustable?

Remains the question: If I have data, and do some math (or have others do the math, i.e. tools and service providers to interpret and analyze the copious amounts of data with electrochemical data models), are the results better and the OEM will accept? Why do results of e.g. SOH assessments differ in the first place?

There is too little standardization on how which KPI is actually defined, and this leaves everybody plenty of room for interpretation, and the OEMs are often defining crucial, warranty-related KPIs like “SOH” themselves in the warranty contracts. All too often however, even this definition is missing or sparse; needless to say that it is crucial to have a zero-ambiguity mutual understanding of key indicators that build the foundation of e.g. warranty contracts.

The main power of independent data analysis is to make a level playing field. Just because there are 2 capable teams on the field now doesn’t automatically mean that one wins. You still have to play (that is: To argue your case towards the OEM), but now one is equipped with a trustable, transparently comprehensible and relatable assessment of the situation. Something that is absent if one must trust the OEMs self-assessment defenselessly. The consequences of differing KPI definitions become very tangible when it comes to capacity testing.

Manual Capacity Tests vs. Real-World SOH from Battery Data

To make it concrete, let’s take a look at manual capacity checks for SOH assessment (again make sure you have read the previous article on SOH!).

A typical procedure is to drain the battery, for instance by driving in circles and/or full-throttle heating, then letting it balance for a few hours to 1 day – super crucial to get the full capacity out, as we learnt recently! – and then perform a charging under controlled conditions, until the battery is, or reports, “full”.

Quite an effort, but is it worth it? Let’s look at another prototypical example: You see only the operational SOH of a LFP bus for 1.5 months, late last year, estimated from telematics battery data. Typical values of 70-80% in daily operation are visible, quite low for a 1-2 year old bus, but with single points close to 90%, too. The reason for the spread is mainly disbalance and the usage profile dependency, but the point to make here is another: 2 OEM-curated capacity tests in the fashion outlined above were conducted in early October, yielding two times 96% — almost 20% higher than the average before. Mind you, both OEM (not shown) and our analysis (shown) do yield 96%.

LFP bus State of Health comparison showing operational battery data discrepancy — Volytica’s capacity analysis, of an LFP bus over 1.5 months shows ~77% SOH in daily operation versus ~96% in OEM-curated capacity tests, revealing a 20% discrepancy.

Contractually and legally, the 96% are the mark: According to the controlled tests under optimal, non-operational conditions, the battery is almost new and far away from the (here) 70% warranty threshold. But under operational conditions and in real life, the operational capacity made available to the owner is significantly lower, and the results of capacity tests should never be used e.g. for route planning or dispatching without thorough scrutiny and deeper understanding!

Manual capacity tests might yield very optimistic results that are not per-se representative for daily operations, and the outcomes can only be used for operational dispatching with a huge grain of salt.

Why KPIs differ, and how to handle it.

The next table shows a comparison of three sources of SOH (OEMs assessment, a capacity check, and using battery raw data) with regard to accuracy, repeatability, independence and effort. It becomes immediately clear that everything has advantages and disadvantages, and every PTO should take a deliberate decision on how to handle data sources, data analysis and how to stand one’s ground.

This comparison highlights key differences in accuracy, independence, effort, and scalability among three SOH assessment methods: OEM onboard estimates, annual cycle tests, and continuous data analysis:

	OEM	Annual Cycle Test	Continous data analysis (e.g. volytica)
Accuracy* Can you trust it? (BMS/ operational net SOH)	(?)**	− (highly influenced by the way the personnel performes; also, batteries must be balanced)	+ (defined, automated algorithms; typically comes with a trustability indicator, too)
Electrochemical / technical nominal and „degrading“ SOH	Not available	Not available	+
Repeatability If performed/ read-out multiple times, will it be consistent?	(?)**	−− (heavily influenced by the people doing it and by the external conditions/ weather, …)	++
Independent Can you use it against the OEM	−−− (obviously not)	+	+
Costs & Time effort Per asset, is the cost & time effort justifiable?	+ (if it can be read out without additional hardware, its certainly cheap)	−− (very labor intensive; vehicle must be driven empty, charging at low power; manual analysis in Excel?)	++
Scalability Can you perform it for your whole fleet	+	−−− (not possible for every vehicle; very labor intensive)	+++
Frequency How often can it be performed to derive actions	+	−−− (max. 1-2/yr)	+

* There are no standards that define „accuracy“ for SOH; also, the „operational / net“ (software-defined) must be distinguished from „technical / nominal“ SOH (electrochemical)

** unclear how the OEMs own assessment should be assess it self, in the first place; fact: the limited computational power of the BMS is very often not able to do accurate / sophisticated algo computations

When it comes to resell value, things are not too different. Instead of PTO discussing with OEM, two PTOs sit on the negotiation table. And again, transparency builds trust, and this trust enables an educated discussion on a fair market price.

Battery Data Transparency at a Glance

Battery data is plentiful, but insight depends on structure, resolution, and trustability, not on data volume alone.
State of Health is not a physical fact but an interpretation that strongly depends on available data, assumptions, and processing methods.
Regulations such as the EU Data Act improve formal access rights, but real-world data access and usability remain primarily a contractual and technical issue.
Manual capacity tests and single SOH values provide only a limited and often optimistic view of real-world battery performance.
Continuous, independent analysis of operational battery data enables a more realistic understanding of residual capacity, degradation behavior, and asset value.

Hands-on Tip #8: Making Battery Data Work in Practice

The EU Data Act is nice, but still it is the buyer to enforce data access, unless you have a trigger-happy, well-financed legal department.
Data quality is crucial; even if you believe you have access to it in the vehicles, make sure it is not just “there”, but in the right resolution, frequency and quality (e.g. check against VDV238 standard).
Also take care that you employ a telematics unit that doesn’t spoil the nice raw data during Cloud-transmission e.g. by applying heavy & lossy compression (what budget telematic providers often do).
Don’t view a third-party / OEM-independent assessment of any performance indicator (e.g. SOH) as the “silver bullet”, yet as the enablement of an educated discussion and a level playing field. Results will differ!