Fleet Software Data Exports: Preparing Electric Fleet Data for Further Use
A good data export from fleet software provides the information needed in a documented, machine-readable structure. For electric fleets, unique vehicle and session identifiers, periods, energy quantities, amounts and assignments are particularly important. Whether an export is useful in practice only becomes clear when it is fully imported and checked against the receiving process's requirements.
A downloadable file therefore answers only part of the question. You also need field meanings, a clear definition of the dataset's scope and rules for subsequent changes. These agreements reduce questions at month-end and prevent important relationships from being lost when data is transferred.
Define the purpose and scope of the export first
Decide whether you need raw data, verified individual sessions or an aggregated analysis. Totals may be sufficient for a cost-centre overview. To investigate an incorrect assignment, however, you need the individual sessions and their identifiers.
Describe the scope independently of the file format. It includes selected vehicles, sites, periods and statuses. Specify whether the period is filtered by session start, session end or another relevant date. A charging session spanning midnight must not appear twice or disappear because of unclear boundaries.
Also define which personal information is required for the purpose. A cost-centre report does not automatically need names or individual driving routes. The export should fulfil its agreed purpose and be restricted to the appropriate recipients.
Create a data dictionary
The data dictionary explains every column's meaning, data type, unit and permitted states. Fields whose names sound familiar but allow several interpretations are especially important.
Field | Meaning to document |
|---|---|
Vehicle ID | Permanent internal identifier; relationship to the current registration number |
Session ID | Unique identifier for an individual record |
Charging start | Timestamp with an unambiguous time zone or time basis |
Energy quantity | Unit and source of the value |
Amount | Currency, net or gross, and included components |
Cost centre | Assignment valid at the relevant time |
Status | For example, provisional, confirmed or corrected |
Distinguish an empty entry from a genuine zero value. Define how vehicles that have left the fleet and cost centres that are no longer valid appear. An export of past sessions should make their original assignment traceable, even if vehicle master data has since changed.
Check the details of the file format
For a CSV export, agree the delimiter, character encoding, column headings and treatment of quotation marks. The technical description in RFC 4180 explains, among other things, how fields containing commas or line breaks are enclosed in quotation marks. It also notes differences between implementations. Source: CSV format description.
Decimal separators, date representation and time zones must also be unambiguous in your export. A value such as “1,250” can be interpreted in several ways without an agreed number convention. Test identifiers with leading zeros too: they are labels and should not be unintentionally converted into numbers.
The general rules for reliable field values are covered in data quality in fleet software. The export agreement supplements these with the requirements of the receiving process.
Practical example: A second import must not duplicate records
Illustrative working example: You export 150 confirmed charging sessions for one month. After review, the cost centre of one session is corrected. You then download the month's dataset again.
If the target system creates every row as a new session, it will contain duplicates. The stable session identifier makes it clear that these are the same 150 sessions. The corrected record also needs agreed treatment, such as replacement after a version check or an explicitly marked correction.
This example shows why a single test export is not enough. Also test a repeated import, a corrected session and a late submission. Record how the target system handles each case and how the result can be verified.
Checklist for export acceptance
Use a small test dataset with deliberately selected special cases. It should include at least a normal session, missing optional information, a corrected record and an identifier with a leading zero.
Do the record count and total match the selected source dataset?
Do identifiers remain unchanged and unique?
Are special characters and umlauts imported correctly?
Are timestamps and period boundaries unambiguous?
Can empty values still be distinguished from zero?
Is a repeated import processed without duplicates?
Are changes and late submissions traceable?
Can authorised recipients use the export independently?
Document the agreed structure with a version number. If fields are added later or their meaning changes, the people responsible for the receiving process need a clear explanation of the change.
Make exports part of your workflow
Define who creates, checks and receives the export. A regular process also needs deadlines and a fallback procedure. Automatic transmission additionally requires access, error handling and monitoring to be clarified. The article on fleet software interfaces and APIs provides a dedicated checklist.
Discuss with StromNow for fleets which charging and billing data your team needs. A data dictionary and an anonymised sample file make your requirements concrete; available export options should be checked against them.
Frequently asked questions
Is a CSV export enough for further processing?
It can be sufficient when fields, format and correction rules are agreed. What matters is whether the receiving process can use the data completely and without manual reinterpretation.
Which identifier should I use to combine data?
Use a stable, unique identifier for the relevant object. A registration number alone is often insufficient for a lasting assignment when vehicles or registration details change.
How can I tell whether an export is complete?
Check the scope, filters, record count and suitable control totals against the source dataset. Also document the data cutoff and any excluded statuses or sessions.