21. Data-Driven Station Development and Regional Analysis

The key issue in developing charging infrastructure is being able to put the right capacity in the right location under suitable operating conditions. To do this, drivers’ needs, existing infrastructure, the electrical connection, site characteristics, and project economics must be evaluated together.

Heavy station searching in a region can be an important signal. However, the reason for this behavior may be as much inadequate information about existing stations, a shortage of certain connector types or a need to compare prices as a real lack of capacity. Sound planning requires understanding the reason behind the observed behavior.

eBattery's regional analysis approach combines usage signals obtainable from the Web2 app, authorized external sources and field assessments in a shared decision process. This process extends from identifying candidate regions to technical feasibility, from project scenarios to post-operation performance evaluation.

The core aim is to turn a need seen in the digital environment into a station project that is physically feasible and economically explainable.

For projects offered under RWA, this approach also helps the investor understand why a station is being developed in a particular area. The data period used, the analysis method, the assumptions and the actual results are linked to the project history.

21.1 Contribution of Web2 Demand Data to Infrastructure Planning

eBattery's Web2 layer can help in understanding the needs drivers encounter while discovering stations and planning trips. Map searches, filter preferences, route plans and user feedback produce different signals that can be considered in infrastructure planning.

The use of this data is carried out within the framework of defined processing purposes, the necessary legal conditions, data minimization and access controls. Sharing information for one function is not to be interpreted as permitting unlimited use of all personal data in investment analyses.

Different behaviors, different levels of evidence

User behaviors are separated according to the nature of the need they express:

Data or behavior Contribution it can make to the analysis Limit of interpretation
Searching for stations on the map Shows discovery interest in certain regions Not counted on its own as a charging transaction or a confirmed demand
Connector, power or operator filter Helps understand the features of the service being sought No filter results may also be caused by missing data
Creating a route to the station Shows that the user evaluated the option in more depth Does not prove that the trip or the charge took place
Need for a charging stop along the route Helps assess the route's energy needs The planned journey is separated from the completed journey
Verified charging session Shows actual service usage Represents only the session scope that is accessible and verifiable
User feedback Makes access, fault or service problems visible Source, freshness and verification status are assessed

This distinction prevents the same user reviewing a station several times from being counted as several independent customer requests.

Separating platform data from the regional market

The behavior of eBattery users does not automatically represent all electric vehicle users in the region. Greater use of the app in a region may cause it to generate denser data, regardless of that region's actual charging needs.

For this reason, analyses take into account user coverage, observation period and regional usage differences. An increase in the number of searches as the platform grows is not interpreted as meaning that new charging demand arises at exactly the same rate.

Web2 data is supported by accessible vehicle counts, existing station records, regional mobility information and field findings. The retention, derivation and sharing conditions of the sources used are maintained.

From app experience to planning insight

Data-driven planning also separates product improvement from infrastructure needs. If the cause of unsuccessful searches in a region is missing station information, the first intervention may be improving data quality. If the problem is a real shortage of capacity, new stations or a capacity increase may be considered.

This way, user behavior is interpreted before it is turned directly into an investment decision.

21.2 Determining Regional Charging Needs

Regional charging demand requires assessing not only where vehicles are located, but also when, for how long and for what purpose they need charging service.

Regular needs around housing, use in a workplace parking area and charging demand along a long-distance route can give rise to different service models. Several of these needs may coexist in the same area.

Defining the service area

The analysis area is not determined by administrative boundaries alone. A station’s accessibility is also assessed through the road network, driving time, entry direction and detour from the route.

A location that looks close on the map may not be practical for the driver because of road junctions or access restrictions. For this reason, regional analysis separates straight-line distance from real travel conditions.

A candidate station’s service area is considered according to its intended type of use. The access logic of an AC solution within a neighborhood is not the same as that of a DC solution on an intercity route.

Separating types of need

eBattery’s assessment approach distinguishes four core needs:

  • Local and regular need: The recurring need of users who do not have sufficient charging access at home or at work.

  • Destination need: Charging needs that arise during shopping, accommodation, work or events.

  • Route need: The energy need that arises along the route to continue the trip.

  • Seasonal demand: Demand arising from holidays, tourism, events or other seasonal movements.

This distinction affects device selection and the economic scenario. A region that is heavily used during certain periods of the year is not modeled as if it had the same level of use all year round.

Number, capacity and accessibility

The number of stations or sockets alone does not explain whether infrastructure is sufficient. When assessing charging infrastructure, the IEA uses the power capacity offered per vehicle as well as the number of charging points per vehicle. Higher-capacity points can, under suitable conditions, deliver different amounts of service in a given time period.

In eBattery's regional analysis too, the number of stations, technical capacity, access conditions and service quality are considered together. The aim is to understand the infrastructure a driver can actually benefit from.

21.3 Route Density and Charging Search Behavior

Route data helps assess on which routes electric vehicles may encounter a need to charge. However, not all vehicles using the same route need to charge at the same point.

The vehicle’s usable battery capacity, starting charge level, consumption, destination and trip preferences affect where charging is needed.

Planned movement and actual movement

A user creating a route does not show that the trip took place. For this reason, planning data is kept separate from verified usage data.

Realized movement or service data is used only when the relevant access and processing conditions are met. Continuous movement tracking is not assumed for a user who has not turned on location sharing.

Alternative plans made on the same route are also not counted as repeated independent trips.

Linking energy needs to the route

For route density to become meaningful in terms of infrastructure, the energy need along the route is evaluated.

The approximate energy level at which vehicles from the starting points could reach a given region is examined together with the available stop options and safety margins. This assessment is made through vehicle and trip scenarios supported by suitable data.

This allows the relationship between heavy traffic and the need for charging services to be established more accurately. A point that many vehicles pass through may not generate the expected level of charging demand, given the existing alternatives and the trip structure.

Separating the reasons behind search behavior

Repeated searches in an area can have different causes. The user may be looking for a lower price, prefer a particular operator, or be unable to find a socket with enough power.

For this reason, filters, the scope of results and related feedback are reviewed together with the number of searches. A shortage of a particular socket type can emerge as a need independent of the overall number of stations.

Campaigns and content within the app can also affect search behavior. Interest that follows heavy promotion in a region is not interpreted as a lasting increase in demand without taking the campaign effect into account.

21.4 Infrastructure Gaps and Use of Existing Stations

An infrastructure gap does not always mean there are no stations on the map. Existing stations being incompatible with the vehicle, being inaccessible at certain hours or frequently going out of service can also create a service gap for the driver.

eBattery's analysis approach assesses different types of gaps separately:

  • Lack of geographic access.

  • Lack of socket or vehicle compatibility.

  • Insufficient power and service capacity.

  • Operating hours or site access restriction.

  • Reliability and availability issue.

  • Service capacity that falls short at peak hours.

This classification ensures that the solution to be developed fits the nature of the problem.

Accurately measuring current usage

Different indicators carry different meanings in station usage:

Indicator What it evaluates
Service availability The station being operational and accessible during its defined service period
Socket occupancy The socket being reserved for a session during the relevant time interval
Active charging time The time during which energy transfer takes place
Energy sales Energy transferred within the service scope in the relevant period
Completed session Successfully completed charging transactions
Session start success The extent to which start attempts turn into service

A socket being occupied is not the same as continuous energy transfer. A vehicle may remain in the space after charging is complete. Similarly, a device appearing online may not mean the driver can physically reach it.

The period and data scope over which the indicators are calculated are clearly stated. Estimates made from intermittent status snapshots are not presented with the same precision as measurements calculated from complete session records.

No data versus no usage

If no data can be obtained from a station, this is not interpreted as the station being unused or empty. Missing data is kept as a separate status in the analysis.

If usage information is not accessible at third-party stations, the capacity assessment is limited to existing technical records and other verifiable sources. If commercial utilization rates are estimated, the method and assumptions are shown clearly.

Regional impact of new capacity

When evaluating the expected usage of a new station, demand that may shift from existing stations is also taken into account. All demand in the same area is not counted again and independently for each new project.

Planned new stations, announced capacity increases and alternative locations are included in the analysis. This separates the additional service a new project will provide from the redistribution of existing demand.

21.5 Location Suitability and Technical Feasibility

Identifying a regional need does not by itself show that a particular site is suitable for investment. The physical and technical conditions of the candidate location are examined separately.

In site assessment, vehicle access to the site, parking layout, user safety, accessibility, environmental conditions, and maintenance access are considered together.

Basic eligibility conditions

The right to use the site, the ability to establish the required connection, and the ability to deliver the service safely are priority conditions. A high demand score is not used to make up for a fundamental shortcoming in these conditions.

For this reason, the assessment is carried out in two stages:

  1. Checking the basic conditions required for the project to be carried out.

  2. Comparing eligible candidates in terms of service and economic characteristics.

This distinction aims to prevent projects that score highly but cannot be implemented from being prioritized.

Field experience from the user's perspective

For the driver, suitability is also related to the station's visibility, the clarity of directions, ease of entry and exit, and waiting conditions.

A location being close to the main road may not be enough. Access from the opposite direction, a long turnaround distance, parking restrictions or opening hours may affect user preferences.

AC and DC solutions are evaluated together with this experience. The aim is a fit between the user's expected parking time and the charging service offered.

Comparison and decision rationale

Candidate locations are compared on demand, access, connection, cost and service conditions. If scoring is used, the meaning and weights of the criteria are explained.

The result score is a tool that summarizes the assessment. It does not replace a site inspection, technical verification or project approval.

The reliability of the data used in the analysis is also shown separately from location suitability. A location that looks high in potential may have limited data coverage; this remains visible in the assessment result.

21.6 Energy Connection, Access and Installation Conditions

The electrical connection directly affects the station's technical capacity, cost and time to commissioning. For this reason, the connection assessment is carried out in the early stages of the project development process.

The U.S. Department of Energy's AFDC guidance also highlights early coordination with the electric utility in charging infrastructure planning, and review of connection needs and possible infrastructure upgrades. The installation budget covers permits, site, maintenance and operating items along with the equipment.

In eBattery’s approach, this general planning discipline is adapted to the connection and operating conditions in each country.

Current and usable capacity

The total capacity of the electrical connection at the site is distinguished from the capacity that can actually be allocated to charging. Other consumption using the same connection, simultaneous loads and operating conditions are taken into account.

The sum of the devices’ nominal power does not always indicate the power that can be delivered at the same time. If power sharing or load management is used, its effect on the service is stated in the model.

Connection cost and timing

Transformer, panel, cabling, excavation, connection and other required works are included in the budget. Supply and connection lead times are reflected in the schedule for entering operation.

It is taken into account that a connection delay may postpone the start of revenue. Rent, financing or preparation expenses that may arise before installation is complete are also assessed in the relevant scenario.

Site and installation layout

Installation conditions cover safe vehicle movement, cable access, physical protection, routing, communication connectivity, and maintainability.

If a future capacity increase is intended, the space and connection conditions required for it are addressed in the initial assessment. However, possible expansion is not presented as the confirmed capacity of the current project.

The technical feasibility result is supported by assessments from authorized experts and the relevant service providers. Regional analysis is not used in place of these technical approvals.

21.7 Revenue, Expense and Usage Scenarios

The economic assessment of a station is based on examining what outcomes may arise under different usage conditions.

Instead of a single usage or revenue forecast, scenarios with explained assumptions are prepared. In each scenario, demand, technical capacity, service availability, prices and costs are treated consistently with one another.

Setting up the use case

A simplified energy calculation can be based on the following relationship:

Periodic energy sales = Number of completed sessions × Average energy per session

This relationship is a calculation example in which sessions are defined with a similar scope. In the real model, time slots, vehicles, power sharing and different usage groups may be assessed separately.

The table below has been prepared only to illustrate the method:

Illustrative scenario Daily completed sessions Energy per session Period Calculated energy
Lower usage 8 25 kWh 30 days 6,000 kWh
Reference use 12 25 kWh 30 days 9,000 kWh
Higher usage 16 25 kWh 30 days 12,000 kWh

These values are not a usage forecast for any eBattery project. Whether the calculated volume is possible at a real station is checked separately against device capacity, operating hours, session durations and site conditions.

If the number of completed sessions already reflects the effect of interruptions, the same effect is not deducted a second time through the availability factor. The place of assumptions in the calculation is kept clear.

Revenue calculation

The revenue scenario is prepared from energy sales, applied tariffs, discounts, refunds and other related service fees.

If different prices apply at different times or for different user groups, they are assessed separately. The amount collected from the customer, including taxes, is separated from the amount reported as project revenue.

Future price changes are not presented as realized revenue. The date and scope of the tariff assumption used are stated.

Expense calculation

Expenses are examined as items that vary with usage volume and items that are relatively fixed.

Energy supply, payment transaction costs and some service expenses may vary with usage. Rent, certain software and operating services, insurance and some maintenance expenses may have different cost structures.

The difference between the electricity purchased and the energy delivered to vehicles is also taken into account. Losses and the station's own consumption are not ignored in the energy cost calculation.

Power- or capacity-based charges in local tariffs are treated separately from a simple calculation based only on the amount of energy.

Sensitivity and break-even assessment

The model is examined against changes such as lower usage, higher energy costs, delayed commissioning or rising maintenance expenses.

In the operating break-even assessment, the question is at what usage level the contribution remaining from each unit of service can cover fixed operating expenses. This indicator is not used in the same sense as recovering the investment cost or generating net profit distributable to investors.

Installation expenses, reserves, taxes and other terms in the project contract are assessed separately. Scenarios help in understanding possible outcomes; they do not constitute a commitment to returns that will be realized.

21.8 Public Presentation of Regional Analyses

Publicly presenting regional analyses aims to let investors assess the rationale behind a project.

In eBattery's approach, showing only a result or a high score is not considered sufficient. Which region, which period and which data scope the result is based on is also explained.

Contents of the analysis view

A public report or project view includes the following elements:

  • The region and service area under review.

  • The observation period used.

  • Information on the types of data sources and the coverage.

  • General view of the existing infrastructure.

  • The type of need identified.

  • Key characteristics of the candidate location.

  • Scenarios used and main assumptions.

  • Analysis date and version.

  • Disclosures about data sufficiency.

Observed data, calculated indicators and forward-looking estimates are shown separately. The user can tell whether a value is a direct measurement or a model result.

Map and comparison presentation

The meaning of the colors and the density scale used in regional maps is explained. When different maps are compared, it is stated if the same color corresponds to different numeric ranges.

Indicators such as the absolute number of searches and search intensity per user are not confused with each other. High app usage alone is prevented from appearing as a large infrastructure gap.

Accessibility for the investor

The regional analysis of an RWA project is linked to the relevant project record. The investor can review separately the report used at the initial presentation and the updates published later.

The same analysis history is also accessible in the secondary market. Past assessments are not hidden by simply replacing them with a more current and more favorable report.

Public disclosure does not mean publishing personal data or third-party data that there is no right to share. The presentation is prepared with information that is permitted and has passed the necessary privacy checks.

21.9 Onchain Recording and Version History of Analyses

Linking regional analysis to an onchain record supports checking afterwards which content and method a given report was published with.

This structure alone does not prove the accuracy of the report. Its main contribution is record integrity and the traceability of the relationship between versions.

Scope of the analysis record

In the on-chain reference structure, the following information can be linked:

Field Function
Analysis ID Identifies the relevant work
Region and scope version Defines which geographic area is being examined
Observation period Shows the time range of the data used in the analysis
Report summary Supports checking the integrity of the published file
Method and model version States the calculation approach used
Assumption version Links the scenario inputs
Publication record Shows which transaction links the report to the chain
Previous version reference Preserves the relationship between updates

The time the data was observed, the time the report was prepared and the time it was recorded on chain are kept separate. The date of the chain record is not presented as a time proving that the physical events took place at that moment.

Updates and corrections

When new data arrives or the method changes, a new version of the analysis is published. The reason for the change and its effect on the result are explained.

If an error is detected in a report, a correction record is created. The validity status of the previous version is stated and its link to the new version is preserved. This ensures that record integrity and use of the correct information are maintained together.

Reproducibility

To be able to recalculate the analysis, a summary of the result file alone is not enough. Under appropriate access, the data scope, calculation rules, parameters and versions must also be preserved.

A public user can check the integrity of a published report. Detailed recalculation, on the other hand, is carried out with access in line with the data-sharing rights and privacy conditions.

Raw personal movement data is not moved on-chain, to preserve reproducibility concerns.

21.10 Forecast vs. Actual Comparison

When the station is put into operation, the estimates used during project development need to be compared with the actual results. This comparison supports both investor disclosure and the quality of evaluation of later projects.

Preserving the initial estimate

The forecast version used at the time of a financing or investment decision is preserved together with the decision date. Updated forecasts prepared later are shown as separate versions.

This prevents the earlier forecast from being changed retroactively in light of actual results. The initial expectation, the current expectation and the actual activity can be reviewed together.

Comparable periods

The first month of operation may not be under the same conditions as a month of full-capacity, full-period operation. Commissioning time, seasonality, maintenance downtime and capacity changes are included in the comparison.

Estimates and realized results are handled on the same period, currency and calculation scope. An estimate including tax is not compared with a realized result excluding tax, nor total collections with net revenue.

Reasons for the deviation

The actual result may differ from the forecast for various reasons:

  • User demand that differs from what was expected.

  • Changes in the amount of energy per session.

  • Tariff and discount effects.

  • Changes in electricity and operating costs.

  • The station’s service availability.

  • New competitor capacity.

  • Delays in installation or connection.

  • Data source or measurement issues.

These causes are separated as far as possible. For example, low energy sales and low collections are not assumed to stem from the same problem.

Model development

Forecast quality is monitored separately for different regions and station types. It is assessed whether the model consistently over- or under-predicts.

In tests run on historical data, information not known at the forecast date is prevented from leaking into the model's input. The data used to develop the model is kept separate from the periods in which performance is evaluated.

The opening of a new station, its visibility within eBattery and campaigns can change demand. Increased usage after opening is not interpreted as confirming all the initial assumptions; the effect of the interventions made is also assessed.

21.11 Anonymization and Data Quality

Regional analysis aims to understand infrastructure needs without revealing users' individual movements. For this reason, the processing, aggregation and public presentation of personal data are carried out with separate control stages.

Scope of anonymization

Removing a username or replacing an identity with another code does not by itself make data anonymous. Location and time patterns, when combined with other information, may make it possible to identify a person.

In KVKK’s explanation, anonymization requires that the data cannot be linked to an identified or identifiable person even when appropriate matching and reversal methods are used.

eBattery’s approach is based on evaluating the data to be published for this linkage risk.

Regional and temporal aggregation

In public analyses, regional and periodic summaries of suitable size are used instead of personal route traces.

Results based on a small number of people may be combined with a wider region or time range, or may not be published. Also taken into account is the possibility that comparing frequently repeated queries or reports close to one another could reveal the behavior of an individual user.

Privacy controls are not applied only to the first report. The information exposure that new versions might create together with earlier versions is assessed.

The core dimensions of data quality

Analysis data is checked from the following angles:

  • Accuracy: The record’s consistency with the event it represents.

  • Freshness: The information being recent enough for the relevant decision period.

  • Completeness: Covering the required fields and observation periods.

  • Consistency: Preserving the same meaning across sources and periods.

  • Uniqueness: Repeated records not being counted as independent events.

  • Representative strength: The extent to which the data reflects the region under review and the user’s needs.

Incorrect coordinates, duplicate events, unit differences and missing time information are handled before analysis. Time zone and period boundaries are applied consistently.

Missing records are not filled in as zero usage. If estimated completion is used, the method is stated and its effect on the calculation is assessed.

Managing artificial activity and conflicts of interest

Automated requests, test accounts, unusual activity driven by campaigns and repeated artificial behavior can affect demand measurement.

Search or engagement intensity that may arise, especially with the aim of raising a region’s investment priority, is examined by comparing it with other sources. A single type of behavior is not allowed to determine a project decision.

Before outliers are simply deleted, it is assessed whether they represent a real surge in demand or a data problem. Not wrongly excluding a genuine increase in demand is also part of data quality.

Making analysis adequacy visible

Not every region may have the same data depth. eBattery shows the scope of the observation used and its level of verification together with the analysis result.

When data coverage is limited, the decision is supported by additional field review and independent sources. The reliability of regional planning rests on correctly understanding the limits of the data used, as well as its quantity, and on continuously comparing the results with operating experience.

Search eBattery

Search for a feature, guide or topic.

Starting pointsFinding a charging station →Sample RWA project →Guides and news →