24. Oracle Infrastructure and Onchain Data Verification

In eBattery's RWA model, the value of the information presented to the investor rests on its verifiable link to physical activity. A station's energy sales, collections, maintenance expenses and distributable net profit must be explained with records whose source, period and calculation method are clear. The oracle infrastructure is the core layer that carries these records from operational systems to applications on the blockchain.

Chargers, payment institutions and accounting systems operate outside the blockchain. Smart contracts, in turn, cannot see on their own the events that take place in these systems. To learn how many kilowatt-hours of energy a station supplied in a given period, or whether a maintenance invoice has been paid, they need data coming from the outside world.

eBattery does not treat this connection merely as a data transfer. It designs together the steps of verifying the source's identity, comparing records, applying calculation rules, checking approvals and publishing the result with appropriate evidence.

The purpose of the oracle infrastructure is to ensure that activities taking place in the physical world are represented in the digital investment system in a consistent, traceable and auditable way. For this purpose, Chainlink-based workflows work together with the station e-ID system, the operational data infrastructure, financial reconciliation mechanisms and onchain reporting contracts.

24.1 The Role of the Oracle Layer

An oracle is the infrastructure that allows smart contracts to use information from outside the blockchain. Within eBattery, this role covers verifying, under certain rules, the data describing station activity and linking it to the relevant project record.

For example, to report monthly energy sales onchain, it is not enough to transfer only the total value visible in the central system. It must be determined which sessions make up the total, whether any records are missing, whether the same session has been counted twice and which time interval was used.

The oracle layer answers five core questions in this process:

  • Which source did the data come from?

  • Which station and operating period does it belong to?

  • Which verification and calculation rules did it pass?

  • Which version and approval status was it published with?

  • For what purposes can the smart contract use this data?

The answers to these questions form the context that travels with the data. The same amount of energy can mean different things in a preliminary report, a period record with completed reconciliation, or a corrected report.

In eBattery’s approach, the oracle does not remove the responsibilities of the source systems. Meter accuracy depends on the physical measurement infrastructure; the accuracy of a payment record depends on the relevant financial source; and the eligibility of an expense depends on documentation and approval processes. The oracle, in turn, ensures that records coming from these sources are processed and published in line with the defined workflow.

Therefore, reliability is not attributed to a single technology. The reliability of the source, the accuracy of processing and the integrity of the published record are assessed together.

In eBattery’s oracle design, Chainlink is positioned as the core component of verifiable workflows between external systems and smart contracts. At the center of this approach is Chainlink Runtime Environment — CRE is found.

CRE makes it possible to combine the steps of pulling data from APIs, reading blockchain records, running calculations and passing results to onchain or offchain systems into shared workflows. Production workflows run on Chainlink’s Decentralized Oracle Networks (DON). The network’s consensus mechanisms support producing a shared result for the relevant operations of the workflow.

eBattery's new development approach is built on CRE. Chainlink presenting CRE as the new direction for its Functions service is also consistent with this choice. In this way, the architecture is based on current workflow infrastructure rather than remaining tied to the way an older product is used.

The workflows designed for eBattery fall into three main groups:

Operational reporting workflows, processes completed sessions, verified energy amounts, data coverage status and station performance indicators.

Financial reporting workflows, checks the period revenues, expense categories, reconciliation results and required approvals, and prepares the report package to be published.

Incident and change workflows, links specific events such as device renewal, report correction or a change in verification status to the project history.

The sources each workflow accepts, the rules it applies and the results it can produce are limited. A workflow authorized to publish energy measurements is not, on its own, given the authority to start an investor profit distribution. The transition between an operational record and a financial result depends on additional reconciliation and approval conditions.

At the move to production, the supported network, service capacity, access conditions and security configurations are verified. Deploying a workflow to CRE's production environment requires access approval; therefore, in eBattery's deployment plan, service access and acceptance tests on the live network are a separate stage.

This architectural choice does not imply a commercial partnership with Chainlink or an endorsement of eBattery's financial reports. Chainlink provides the technical infrastructure of the defined data workflow; the content of project data and operational responsibility remain with the relevant sources and eBattery's control processes.

24.3 Operational and Financial Data Sources

The quality of the oracle infrastructure is directly related to the nature of the sources it accesses. eBattery defines the authoritative source and the verification method separately for each data type.

It is the appropriate source metering system for the energy amount of a charging session. For a payment to take place, the payment provider’s records are needed. The scope of a maintenance service, in turn, is explained by the work order, service document, and invoice. Using these sources interchangeably weakens the meaning of reporting.

Source group Information it provides Basic control approach
Charger and meter Energy measurements, session events, device status Device identity, measurement sequence and integrity checks
Charging management system Session relationships, tariff references, operational status Deduplication, period and record completeness
Payment service provider Authorization, collection, refund and settlement Transaction reference and session matching
Bank and financial records Funds movements that have taken place Reconciliation of account, amount, currency and date
Energy supply documents Energy cost and related billing items Consumption period and metering boundary check
Accounting and document system Revenue, expense, tax and period records Document linkage, classification and approval
Maintenance and site management Fault, response, part replacement and service acceptance Work order, acceptance record and expense link

For each source, access authorization, update frequency and responsible party are defined. Where a source does not provide an automatic API, controlled document-based data entry may be used. However, such a record is not presented under the same method label as automatic data taken directly from the source system.

Document reading and classification tools can speed up processing. However, an amount extracted automatically from an invoice is not accepted as a final expense until it is matched with the document and passes the required checks. AI-assisted processing is subject to the same principle: a suggested classification is separate from an approved financial record.

The source chain is preserved up to the aggregation presented to the investor. This way, which records make up an expense category in the report can be traced back within the scope of an authorized review.

24.4 Multi-Source Reconciliation and Consistency Checks

Multi-source verification is broader than fetching the same information again from different servers. If two APIs are fed from the same database, their results matching does not constitute independent verification.

eBattery also records dependencies between sources. Device measurement, payment provider records and bank settlement can complement each other because they observe different events. Two reports that merely copy the same station total, however, do not provide a new layer of evidence.

Reconciliation checks are applied mainly in the following areas:

Identity consistency: The data is checked to confirm that it belongs to the correct station, device, session and project.

Time consistency: Measurements and financial records are assessed with the same period boundaries. Different time zones and late-arriving events are taken into account.

Quantity consistency: The consistency of energy, fee and payment amounts with the relevant rules is reviewed.

State consistency: Transactions that were cancelled, failed or carried out for testing are not mixed with normal commercial activity.

Document consistency: Processing the same invoice more than once, or repeating the same expense in different categories, is prevented.

Some of the checks require an exact match. Two records with a different project ID or currency are not accepted as approximately similar. For physical measurements, tolerances justified by the measurement limit and equipment specifications may be used. Tolerances do not create a general margin of freedom that would make differences invisible.

For example, the energy measured at the site entrance may not be expected to equal the total energy delivered to vehicles. There may be auxiliary consumption and conversion losses. However, if the difference goes outside the normal range, it requires review in terms of the meter, time matching, missing sessions or physical equipment.

Financial discrepancies are not resolved by averaging. If the payment provider shows 1,000 units collected and the accounting record shows 900 units of revenue, the oracle does not treat the midpoint of these two amounts as correct. Tax, refunds, deductions, period differences or recording errors are investigated.

The goal of the multi-source approach is to ensure that different records correctly explain the same economic event, before agreeing on a single figure.

24.5 Data Signing, Timestamps and Evidence Records

eBattery’s proof structure consists of source signatures, report integrity digests, timestamps, and onchain transaction references. Each of these components answers a different question.

Digital signature, helps check that the record was approved by a specific key and whether it was changed after being signed.

Hash — cryptographic digest, is used to compare the integrity of a specific data or file version.

Timestamp, shows when the event was reported to have occurred, when the data was received and when the result was published.

Onchain transaction reference, states in which blockchain transaction the report was recorded.

Signatures are separated by their roles. The device signature represents the measurement record; the authorized financial approval represents the relevant control stage of the report; and the oracle report signature represents the result produced by the workflow. None of these signatures replaces another.

The report package includes, at a minimum, the following context:

  • Station e-ID and report ID.

  • Operating period and report type.

  • Data schema and calculation method version.

  • Integrity digest of the source record package.

  • Verification and approval status.

  • The report version and, if any, its relationship to the previous version.

  • Target network and recipient contract context.

  • Sequence or unique transaction information that prevents repeated processing.

The data format is also defined in hash generation. If field order, numeric precision, currency display and date format are not consistent, records with the same meaning can produce different digests. For this reason, calculations and proof packages use a versioned representation standard.

The time of recording on the blockchain does not prove that the physical event happened at the same moment. For example, an energy report published at month-end may include the previous month's activity. Event time and onchain publication time are shown separately.

Similarly, the presence of a document's hash on the chain does not by itself prove that the document's content is correct or that the expense is commercially appropriate. Proof of integrity gains meaning together with document review and authorized approval.

24.6 Onchain Transfer of Operational Data

Charging stations produce large amounts of technical data. Turning every instantaneous power measurement or device message into a separate blockchain transaction can create cost and complexity that investors do not need.

eBattery summarizes operational data according to its purpose of use. While detailed source records are kept in the appropriate archive, the onchain layer holds records that describe project performance and can be linked to their source when needed.

Operational reports may cover the following information:

  • Number of commercial sessions completed in the relevant period.

  • The verified amount of energy.

  • Service availability and measurement method.

  • Summary of outage and maintenance durations.

  • Data coverage rate and missing record status.

  • Corrections carried over from the previous report.

  • Integrity reference of the source data package.

The scope of each total is defined. For example, test sessions are separated from the number of commercial sessions. A time period with missing data is not shown as zero activity.

Proof methods such as a Merkle tree can be used for batch records. In this structure, a shared root digest of many records is kept onchain; whether a specific record is included in the relevant batch can be checked with a separate proof. Such a proof shows that the record is included in the batch. It does not separately prove that the batch covers all activities completely or that the content is physically accurate.

In onchain transfers, the authorization of the receiving contract also matters. Chainlink’s consumer contract approach offers checks aimed at verifying the expected forwarder contract and the appropriate workflow ID. In eBattery’s design, project, report type, version, target network and replay-protection checks are added to these.

When a transaction is submitted, the report is not considered final immediately. The states of being submitted to the network, being accepted by the contract, and being completed under the relevant network's finality policy are tracked separately.

24.7 Onchain Recording of Revenue and Expense Reports

eBattery's approach to financial transparency requires disclosing not only station revenue but also the expenses incurred to generate that revenue.

An onchain financial report does not consist of a single “profit” amount. A periodic record structure is used that shows how revenues, costs, reserves and the distribution decision are linked to one another.

Report component Disclosed content
Operating revenues Revenues arising from the charging service and recorded for the relevant period
Collection status Amounts collected, pending, refunded or disputed
Operating expenses Energy, rent, maintenance, insurance, payment and other eligible expenses
Operating service fee The operating fee of eBattery disclosed in the project terms
Taxes and obligations Items taken into account in the calculation for the relevant period
Reserve movements Amounts set aside or used for major maintenance and equipment renewal
Period adjustments Disclosed corrections linked to earlier records
Distribution account The amount found eligible for distribution under the project rules

The accounting result, cash position and distributable amount are shown separately. A revenue entry having been recorded may not mean the related money has been collected. Likewise, setting aside a reserve and actual maintenance spending are different transactions; calculation rules are applied that prevent the same economic cost from being deducted twice.

Project financing raised from the RWA primary sale is not reported as charging operating revenue. The price paid to the seller in a resale of an investment unit is also not the station’s operating revenue. In this way, investment transactions are kept separate from the economic performance of the physical service.

eBattery's operating service fee is tracked clearly among the expenses. The platform commission on profit distribution being 0% does not mean the operating service is free of charge. The basis, period and calculation method of the relevant fee are shown consistently with the project documents.

Recording a financial report onchain and making a payment to an investor are separate transactions. The distribution contract applies additional checks on the valid report version, authorized approvals, entitlement rules, available funds and payments previously made for the same period.

A report having been published does not by itself constitute a distribution order.

24.8 Relationship Between Onchain Records and the Document Archive

The onchain record preserves the report’s integrity and version history, while the document archive stores the detailed supporting evidence. The two layers complement each other.

The archive may contain meter record packages, payment settlements, invoices, maintenance documents, contract references and approval records. Each document is linked to the relevant project, period and report version.

Document access is arranged at three levels:

Public information, includes the station's activity summaries, disclosed revenue and expense categories, report versions and verification status.

Authorized investor informationcovers the details required by the investment relationship and the applicable access rules.

Restricted documents, personal data, security information, trade secrets and records that only assigned teams or authorized reviewers can access.

Transparency does not require drivers' personal information or all identifiers on invoices to be made public. Appropriate masking and summarization are applied to publicly available documents. It is kept clear that the original document and the public version are different files and may have different integrity digests.

Using a hash also does not automatically anonymize personal data. Especially for content that can be guessed or linked with other data, the risk of re-identification is assessed separately. For this reason, onchain proof packages are designed with data minimization.

The accessibility of the archive is a separate operational responsibility. A file's hash being on the blockchain does not ensure the file remains accessible in the future. Backup, access management, protection of encryption keys and restore checks are part of the document infrastructure.

This way, the record presented to the investor does not remain just a number published in the past; it keeps its link to supporting evidence that can be reviewed at the appropriate authorization level.

24.9 Data Freshness and Verification Status

It matters not only that a piece of information is correct but also which point in time it represents. The station status from the last five minutes and a closed financial report for the previous month are not assessed by the same freshness standard.

eBattery defines an update and validity policy for each data class. Operational data is refreshed more often, while financial reports depend on period close and the reconciliation process.

The verification status is not limited to a single "approved" label either:

Status Meaning
Received Source data has reached the system
Passed technical review Format, identity and basic integrity checks completed
Awaiting reconciliation Matching with other related records is in progress
Reconciliation completed Defined consistency checks passed
Authorized approvals completed Approvals required by the report type were obtained
Published onchain Record written to the relevant contract
Finalized The network's defined finality condition was met
During an objection or correction process There is an open review on the record
Replaced by a new version The current valid record has been moved to another version

In addition to these, the data coverage ratio is shown. A total produced from records covering only part of the period is not presented as a full-period result.

For data that exceeds the freshness period, the last known value may be retained; however, the current-data label is removed. Not receiving new data does not automatically mean zero production or zero revenue.

Smart contracts also apply validity checks appropriate to the report type. An old report is not allowed to be reused for a new financial transaction. A historical report, on the other hand, remains accessible for the purpose of explaining a past period.

When required data or approval is missing, the relevant automated process is put on hold. The system does not produce a verified result by silently filling in the missing information.

24.10 Correction, Objection and Re-approval Process

Physical operations and financial records may need correction over time. A late invoice, a payment refund, a meter-matching error or a wrong expense classification can affect a previous report.

In eBattery’s approach, a correction is not made by erasing history. The old record is preserved; the new record references the previous version and explains the reason for the change.

The correction process consists of the following stages:

  1. A dispute or error record is created. The affected report, period and line item are identified.

  2. Evidence is collected. Source records and related documents are compared.

  3. The impact is assessed. It is determined whether the issue affects the operational indicator, the financial result or the distribution.

  4. A new report is prepared. The calculation is run again under the same defined rules.

  5. The required approvals are obtained. The roles of the person who prepares a correction and the person who approves it are separated.

  6. A new version is released. The link to the previous version and the reason for the change are recorded.

  7. Linked processes are updated. App screens and, if needed, subsequent distribution accounts are linked to the new status.

Opening an objection does not mean the record is automatically wrong. That said, a dispute with a high economic impact may require the related distribution to be held until the review is complete. The scope of the hold is limited to the project and transaction affected by the issue.

A payment that has already been made is not considered reversed by changing the report version. Differences that emerge after distribution are handled according to the correction and netting rules in the project contracts. Collecting funds from an investor's wallet unilaterally and without explanation is not part of the data correction mechanism.

Protection against repeated processing continues here as well. Resubmitting the same correction report does not create a second correction or payment effect.

The investor can see the original version of the relevant report, the later version and the reason for the change together. This approach preserves accountability as well as accuracy.

24.11 Oracle Trust Model and Control Layers

eBattery’s oracle trust model consists of control layers with different responsibilities. Each layer reduces specific risks; none of them takes on all of the others’ responsibilities.

Control layer Core responsibility
Physical measurement Reliable generation of meter, device, and field data
Source ID Verifying the system that produced the record and its authority
Data processing Deduplication, period matching and calculation rules
Multi-source reconciliation Comparing operational and financial records
Authorized approval Approval of the report under commercial and financial responsibility
Oracle execution Running the defined workflow and producing the report
Smart contract Checks for authority, version, validity and reuse
Monitoring and review Tracking of deviations, outages and authority changes

The Chainlink DON structure provides consensus aimed at reducing the dependence of oracle operations on a single executor. However, when multiple nodes read the same faulty source data, adding more nodes does not by itself correct an error in the physical source. For this reason, oracle node diversity and data source independence are assessed as separate trust properties. CRE’s production settlement also differs from the local simulation model; a successful result in simulation does not mean that multi-node operation and security controls have been tested in production.

In eBattery's production acceptance process, scenarios such as forged source responses, missing records, duplicate submissions, expired reports, unauthorized workflows and connection interruptions are tested. In addition to producing correct results under normal conditions, these tests aim to verify safe behavior under faulty conditions.

Administrative authorities are also protected separately. Transactions that change the source list, verification thresholds, workflow version or report acceptance rules are recorded. Significant changes affecting the economic result are carried out with segregation of duties and suitable multi-approval mechanisms.

When oracle access is interrupted or the required reconciliation cannot be achieved, financial automation is paused. The interruption does not disable the charging devices’ physical safety and local operating mechanisms. During recovery, operation continues from the last valid state; pending reports are not finalized in bulk without being rechecked.

This trust model makes the scope of every significant piece of data eBattery presents to investors explainable: which source it came from, who approved it, which rules processed it and which record published it can be traced. The Oracle infrastructure thus combines the physical activity of the charging station with the transparency and accountability that the investment relationship requires.

Search eBattery

Search for a feature, guide or topic.

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