Value Stream Mapping for Tolling and Road User Charging: From Trip Detection to Billed Journey Without the Dispute Loop

In tolling and road user charging, the value stream is unusual: almost everything flows as information rather than physical material. A vehicle passes a gantry, a detection is created, data is matched to a tag or plate, a tariff is applied, and payment is posted.

The customer may experience only one visible outcome, a correct charge, but the back office may contain hidden queues, repeated validations, image-review work, delayed account lookups and reconciliation breaks. These queues are invisible in the system architecture but expensive in customer experience, revenue assurance and operating cost.

Value Stream Mapping (VSM) makes that hidden flow visible. It connects roadside detection, transaction processing, billing, payment and exception handling into one end-to-end view. The result is a practical basis for removing waste, reducing variation and improving the journey from trip detection to cash collected.

The worked example below is an illustrative operating model for a free-flow tolling and road user charging service.

1. Scope Selection: Define the Product Family and Takt Time

A useful VSM begins with a clearly defined product family. For this map, the product family is:

• Tag-account trips: trips matched to a valid electronic toll collection account.
• Video-matched plate trips: trips identified through automatic number plate recognition and matched to a customer, fleet or registered vehicle.
• Road user charge billing: rated journeys that result in an account debit or invoice.

The map boundaries are:

Start: Gantry detection
End: Cash collected, posted and reconciled

This boundary includes:

1. Gantry detection
2. Transaction ingestion
3. Tag or plate matching
4. Image review
5. Account lookup
6. Trip pricing and business-rule application
7. Invoice or account debit
8. Payment posting
9. Exception and dispute handling
10. Financial reconciliation

Assume the operation processes 120,000 trips per day across a 20-hour processing window.

[
\text{Takt Time} = \frac{72,000 \text{ seconds}}{120,000 \text{ trips}} = 0.6 \text{ seconds per trip}
]

At the broader service level, the operation must absorb approximately 6,000 trips per hour. The transaction-matching platform has capacity for 6,750 trips per hour, creating 12.5% headroom. However, the image-review process does not have equivalent capacity. That imbalance becomes the principal constraint.

2. Current-State Map: Where the Queue Actually Lives

The following table shows the current-state path. Cycle time is active processing time; wait time is queue or handoff time.

Worked current-state numbers

Of the 120,000 daily trips:

• 110,400 trips are auto-matched at 92%.
• 9,600 trips require additional qualification.
• 7,200 trips per day enter manual image review after automated recognition and rules-based checks.
• Review throughput is only 5,800 items per day, creating a net queue increase of 1,400 items per day.
• Average review WIP is approximately 12,600 transactions.
• Approximately 4,300 unmatched transactions are more than 48 hours old.
• The dispute rate is 42 disputes per 10,000 trips, or approximately 504 disputes per day.

The current median time to bill is 2.8 days. Time to cash is 8.6 days. Active value-added work totals approximately 11.6 minutes per trip.

[
\text{Flow Efficiency} = \frac{11.6}{8.6 \times 24 \times 60} \times 100 = 0.09%
]

That is the defining VSM insight: the process is not slow because each transaction requires extensive work. It is slow because information waits between systems, queues and decisions.

The Process Cycle Efficiency Calculator can help teams quantify the same value-added and non-value-added split in their own operation.

3. DOWNTIME Waste in Tolling Operations

Defects

• Incorrect OCR creates a false plate-to-account match.
• Duplicate gantry events produce duplicate charges.
• Incorrect vehicle classification applies the wrong tariff.
• A missing image prevents a valid transaction from being billed.

Overproduction

• Repeatedly generating status reports that no operational owner uses.
• Producing duplicate invoices when a payment file is delayed.
• Creating separate records for the same journey across interoperable agencies.

Waiting

• Transactions wait for manual image review.
• Customer disputes wait for evidence retrieval from another agency.
• Payment files wait for end-of-day bank or card settlement.
• Registration lookups wait for external data responses.

Non-utilised talent

• Skilled analysts spend time correcting avoidable OCR errors.
• Customer-service specialists manually search systems for information already available in transaction data.
• Finance staff investigate routine reconciliation breaks instead of analysing systemic causes.

Transportation

• Transaction files move between multiple platforms and agency hubs.
• Images are transferred between roadside, vendor and review environments.
• Dispute evidence is routed between toll operators, account providers and customer-service teams.

Inventory

• 12,600 transactions sit in the image-review queue.
• Aged unmatched items accumulate because daily inflow exceeds review capacity.
• Unresolved disputes become an inventory of customer-service work.

Motion

• Reviewers switch between image tools, account platforms and tariff systems.
• Agents re-enter plate numbers, dates and locations into separate screens.
• Finance teams search across multiple reports to find one missing settlement record.

Excess processing

• Multiple people approve low-risk adjustments.
• The same image is reviewed after automated recognition, manual review and dispute escalation.
• Reconciliation teams compare files manually even when a transaction identifier exists.

The three largest quantified opportunities are:

1. Manual image review:
   7,200 transactions × 90 seconds = 180 staff-hours per day.

2. Dispute handling:
   504 disputes × 18 minutes = 151 staff-hours per day, before supervisor review.

3. Aged unmatched transactions:
   A queue growing by 1,400 items per day creates a compounding customer-service, revenue-assurance and reporting burden.

4. Future-State Build: Design for Flow and Built-In Quality

The future-state map should reduce human intervention without weakening financial control.

Each transaction should carry one immutable identifier from detection through to ledger posting. That identifier should connect the roadside event, evidence image, match decision, trip, tariff, invoice, payment, dispute and adjustment.

This creates built-in traceability, reduces repeated searching and allows exceptions to be managed as a controlled flow rather than as disconnected cases.

5. Current Versus Future-State Data

These targets are not achieved by simply asking reviewers to work faster. The essential design change is to reduce the volume entering the queue, resolve low-risk exceptions automatically and reserve specialist attention for genuinely uncertain transactions.

The FHWA’s overview of electronic toll collection describes the importance of back-office transaction processing, image review, customer account management, payment settlement and reconciliation. A practical external comparison is the A24 free-flow tolling case study, which reported 96.3% automatic qualification and 3.7% manual image review across a complex multi-jurisdiction environment.

6. 90-Day Kaizen Sequencing

Wave 1: Days 1–30, Stabilise and Measure

Owners: Toll operations manager, data analyst and finance-control lead

Actions:

1. Establish the transaction-level identifier across all process steps.
2. Measure auto-match, image-review arrivals, throughput, WIP, ageing and disputes daily.
3. Create a Pareto of unmatched causes: unreadable plate, missing image, conflicting tag, invalid account, duplicate event and tariff error.
4. Introduce a FIFO review lane with a 72-hour ageing limit.

Metrics:

• Baseline accuracy by cause
• Review queue WIP
• Median time to bill
• Disputes per 10,000 trips
• Reconciliation breaks by source

Expected result: Stop queue growth, create reliable baseline data and expose the three highest-impact causes.

Wave 2: Days 31–60, Improve the Constraint

Owners: Product owner, ANPR or data-science lead and customer-service manager

Actions:

1. Introduce confidence scoring using tag, plate, time, location and account signals.
2. Pilot dark processing for low-risk exceptions.
3. Standardise image-review work using decision rules and a single evidence screen.
4. Launch self-service dispute evidence for duplicate and “not my vehicle” claims.

Metrics:

• Auto-match rate
• Manual review hours per 10,000 trips
• First-time dispute resolution
• Review accuracy
• Percentage of disputes resolved through self-service

Expected result: Increase auto-match to 96%, reduce manual review effort by at least 30% and cut dispute cycle time below three days.

Wave 3: Days 61–90, Flow, Control and Scale

Owners: Operations director, finance controller and technology delivery lead

Actions:

1. Implement straight-through billing for approved confidence bands.
2. Connect transaction, payment and ledger reconciliation through the immutable event ID.
3. Deploy a 15-minute performance dashboard.
4. Set daily management routines for ageing, exception ownership and revenue leakage.
5. Document standard work and control limits for key process measures.

Metrics:

• Time to bill
• Time to cash
• Cost per trip
• Exceptions per 10,000 trips
• Reconciliation break rate
• Aged WIP
• Revenue recovered from previously unmatched transactions

Expected result: Reach the future-state targets, sustain the gain and create a repeatable improvement system for new tolling schemes and road user charging products.

Build the Capability to Improve Transaction Services

Value Stream Mapping is especially powerful in tolling because the most expensive waste is often hidden inside digital handoffs, queues and exception loops. A well-designed map reveals where customer experience, revenue assurance and operating cost intersect.

Build the capability to lead this work through CSSC-accredited Lean Six Sigma Green Belt and Black Belt training at lean6sigmahub.com. Learn how to apply DMAIC, VSM, root-cause analysis and statistical decision-making to complex transaction services.

References

• Federal Highway Administration: Nationwide Electronic Toll Collection Interoperability
• Emovis: Powering the Netherlands’ First Free-Flow Tolling System
• Lean Six Sigma Concepts and Glossary
• Process Cycle Efficiency Calculator

Kaizen. Kai-Care. Kai-Done. Lean Six Sigma