Transport safety economics · Scenario analysis
Cost-Effectiveness of Per-Seat Seatbelt Sensors and Alerts on Buses: An Australian Benefit–Cost Analysis
Abstract
Objective. To establish the conditions under which per-seat seatbelt sensors and alerts produce a positive social return on Australian buses. The analysis connects correct belt use, passenger injury risk and the lifetime cost of an installed system.
Methods. A deterministic economic model compares an equipped bus with the same service operating under its existing reminder practices. Published Australian economic guidance supplies the appraisal method. A government assessment of car reminders and a Korean bus pilot supply historical and behavioural context. The worked scenario assigns a $5,000 installation cost, $100 annual maintenance, ten years of service and a 7 per cent real discount rate. Correct belt use increases from an assumed 20 to 60 per cent. An assumed 50 per cent reduction in the modelled injury risk from wearing a belt is applied to a baseline annual passenger trauma burden of $4,500 per bus. All scenario amounts use constant 2025 Australian dollars.
Results. The scenario produces a 22.2 per cent reduction in the modelled trauma burden and $1,000 in annual benefits. Discounted benefits total $7,024 and discounted costs total $5,702. The benefit–cost ratio is 1.23, net present value is $1,321, and equivalent annual net benefit is $188 per bus. Break-even requires $812 in annual benefits, corresponding to a wearing rate of 52.5 per cent under the specified risk assumptions. Ratios range from 0.31 to 1.54 as the assumed wearing rate rises from 30 to 70 per cent.
Conclusion. Economic value depends jointly on additional correct belt use, baseline passenger trauma risk and system cost. The model supplies explicit thresholds for fleet purchasing and a reproducible framework for incorporating Australian trial and casualty data.
Keywords: bus safety; seatbelt reminders; buckle sensing; benefit–cost analysis; injury prevention; threshold analysis.
1 The economic question
A seatbelt warning system creates an economic benefit when its effect on passenger behaviour reduces expected injury. Each installation has an immediate cost. Its safety benefits accumulate over the journeys made during its working life. The decision therefore depends on the value of the additional protection relative to the full cost of providing it.
This study asks three questions. How can additional correct belt use be converted into an expected reduction in passenger trauma? What benefit–cost ratio follows from an explicit set of operating and cost assumptions? What minimum improvement in wearing would justify the investment?
The study uses a social perspective, counting resources used by operators and safety benefits received by passengers and the wider community. The unit of analysis is one belt-equipped bus operating a defined service. Fleet results are obtained by combining buses with their own exposure, costs and operating lives. The analytical contribution is a set of reproducible scenario results and decision thresholds.
2 Evidence supporting the appraisal
2.1 The Australian car-reminder precedent
The 2012 Australian Regulation Impact Statement reported a benefit–cost ratio of 1.25 and an annual net benefit of $276,219 for driver reminders. Its selected case used a 15-year crash life, 7 per cent device effectiveness and 7 per cent discounting. Simple driver reminders cost $18 in 2010 prices. Existing voluntary fitment was 89 per cent.1
The appraisal updated an earlier injury model, valued reductions in fatal, serious and minor injuries, and allowed for existing fitment. Its ratio described the additional national effects of regulation. We adopt the sequence of injury modelling, valuation and discounting for bus passengers, with inputs defined for the bus application.
“The method of analysis used in 2004 is still valid today”Department of Infrastructure and Transport, 2012, p. 7.1
2.2 A bus application of passenger-level sensing
Park, Lee and Jun reported a Korean interurban bus pilot conducted in 2015. Each passenger seat had occupancy and buckle detection, a visual warning and a link to a driver display. The bus carried 45 passenger seats with two-point belts. Observed wearing was 9.6 per cent in the comparison bus and 59.0 per cent in the equipped bus, a difference of 49.4 percentage points. An analysis of recorded occupied-seat observations produced a wearing rate of 55.82 per cent.2
This observational comparison documents a practical chain from sensing to passenger warning and belt use. Its visual-alert design supplies evidence about that configuration. An Australian evaluation can estimate the effect of its specified alert design through a controlled comparison and direct validation of correct wearing.
| Source | Quantity reported | Role |
|---|---|---|
| Australian government car appraisal1 | BCR 1.25; annual net benefit $276,219 | Historical economic precedent |
| Korean observational bus pilot2 | Observed wearing: 9.6% and 59.0% | Behavioural evidence for seat-level visual warnings |
| Present constructed bus scenario | Assigned wearing: 20% and 60% | Inputs for calculating economic thresholds |
The measures in Table 1 answer different questions. The economic ratio relates money benefits to money costs. The wearing rates describe passenger behaviour. The model below connects these quantities through a separate estimate of injury protection.
3 From belt use to economic value
3.1 Define the system and the comparison service
The intervention comprises occupancy sensing, buckle sensing, passenger alerts, a driver interface, installation and ongoing support. Its specification records when alerts begin, how they are presented and when they stop. The comparison service retains the reminders, supervision and equipment already in use. Both cases carry the same passengers over the same routes in the model.
The primary behavioural measure is the share of passenger exposure with a belt correctly worn. Passenger-kilometres provide a useful denominator because longer journeys contribute more exposure. Observers validate the sensor record against belt placement on occupied seats. Wearing rates include periods of faults, bypassing and reduced response so that the estimate represents the performance of the deployed system.
3.2 Calculate the change in injury risk
Let p0 be baseline correct wearing and p1 be wearing with the system. Let e be the proportional reduction in a specified casualty risk when a belt is worn. For an unbelted risk q, average risk is q(1 − pe). Holding crash exposure constant gives the following reduction relative to baseline:
The denominator accounts for the protection already provided by passengers who wear belts under the comparison conditions. Applying the reduction to baseline casualties gives the number prevented. The calculation is performed separately for each crash type and casualty outcome where their risks or belt effects differ.
Bt = Σj(ΔNj,t × Vj)Equation 2. Annual avoided casualties and their monetary benefit.
Here, Nj,t is the expected baseline count for casualty category j in year t, and Vj is its economic value. Categories are mutually exclusive. A change from fatal injury to survival with injury is represented through the resulting outcome counts. Risk estimates refer to the eligible bus passengers and journeys receiving the intervention.
The scenario model assumes equal underlying crash exposure for passengers whose wearing changes and other passengers in the same group. In a fleet appraisal, route, service, belt type and passenger differences are represented by separate groups. Each group receives its own wearing and risk estimates before benefits are added.
3.3 Value outcomes and resources consistently
The Office of Impact Analysis gives a value of statistical life of $5.87 million in 2025 dollars. This values reductions in fatality risk. Its guidance also describes valuing injury through the severity and duration of health loss.3 The present scenario expresses all monetary inputs in constant 2025 Australian dollars.
A detailed appraisal uses a documented valuation for each outcome. Where medical or productivity components are included in a selected unit value, those components are counted once. Operator insurance payments and compensation transfers are recorded separately from the underlying social losses they finance.
Installed cost includes sensors, wiring, interfaces, commissioning and vehicle downtime. Annual cost includes maintenance, subscriptions, fault response and staff time. Replacement equipment is assigned to its expected replacement year. Any remaining asset value is credited at the end of the evaluation period. Research-only measurement activity is recorded in the study budget; routine equipment and work required to deliver the system enter the deployment appraisal.
3.4 Discount the annual flows
We use a 7 per cent real discount rate and examine 3 and 10 per cent, following OIA appraisal guidance.4 Installation occurs at year zero. Annual benefits and operating costs occur at each year-end. The benefit–cost ratio and net present value use the same population, price base and time period.
PV(C) = C0 + Σt=1…T Ct / (1 + r)t
BCR = PV(B) / PV(C)
NPV = PV(B) − PV(C)Equation 3. Discounted economic outcomes.
For constant annual amounts, the annuity factor is A = [1 − (1 + r)−T] / r. Multiplying an annual amount by A gives its present value. Dividing NPV by A gives equivalent annual net benefit. This places the up-front investment and annual safety benefits on a common annual basis.
4 The constructed bus scenario
Table 2 specifies the numerical experiment. The assigned baseline trauma burden is the expected annual monetary loss for the passenger outcomes represented in the model. It combines their frequency and severity. A common belt effect is applied to that burden in this aggregate scenario.
| Parameter | Value | Meaning |
|---|---|---|
| Installed cost, C0 | $5,000 | Complete system fitted to one bus |
| Annual operating cost, M | $100 | Recurring support and maintenance |
| Service life, T | 10 years | Constant annual operation |
| Real discount rate, r | 7% | Central appraisal rate |
| Baseline correct wearing, p0 | 20% | Comparison service |
| System correct wearing, p1 | 60% | 40 percentage-point increase |
| Belt effect on modelled risk, e | 50% | Relative reduction for a belted passenger |
| Annual baseline trauma burden, H | $4,500 | Expected passenger harm per bus-year |
| Replacement and residual value | $0 | Equipment lasts for the evaluation period |
Equation 1 gives (0.60 − 0.20) × 0.50 / (1 − 0.20 × 0.50) = 0.2222. The expected trauma burden therefore falls by 22.2 per cent. Annual benefit is $4,500 × 0.2222 = $1,000. The ten-year annuity factor at 7 per cent is 7.0236.
The analysis varies the achieved wearing rate and discount rate independently. It then solves for the annual benefit, baseline trauma burden and wearing rate that give a BCR of exactly one. All calculations retain full precision; displayed money values are rounded to the nearest dollar.
5 Results and decision thresholds
5.1 Central scenario
| Outcome | Calculation | Result |
|---|---|---|
| Annual benefit | $4,500 × 22.222…% | $1,000 |
| Present-value benefits | $1,000 × 7.0236 | $7,024 |
| Present-value costs | $5,000 + $100 × 7.0236 | $5,702 |
| Benefit–cost ratio | PV(B) / PV(C) | 1.23 |
| Net present value | PV(B) − PV(C) | $1,321 |
| Equivalent annual net benefit | NPV / 7.0236 | $188 |
Each dollar of discounted cost yields approximately $1.23 in discounted social benefit in this scenario. For 100 identical buses installed together, the corresponding net present value is approximately $132,122. Multiplication changes the total benefit and cost by the same factor, leaving the BCR at 1.23.
5.2 The wearing rate determines the return
| Achieved wearing | Increase, percentage points | Annual benefit | BCR |
|---|---|---|---|
| 30% | 10 | $250 | 0.31 |
| 40% | 20 | $500 | 0.62 |
| 50% | 30 | $750 | 0.92 |
| 60% | 40 | $1,000 | 1.23 |
| 70% | 50 | $1,250 | 1.54 |
The same equipment crosses the economic break-even point as its sustained wearing effect increases. This result gives behavioural measurement a direct decision role. A purchaser needs the achieved improvement over the service’s existing practices, together with its cost of delivery.
5.3 Discount-rate sensitivity
| Real rate | BCR | NPV | Annual benefit for break-even |
|---|---|---|---|
| 3% | 1.46 | $2,677 | $686 |
| 7% | 1.23 | $1,321 | $812 |
| 10% | 1.09 | $530 | $914 |
Higher discounting reduces the present value of future safety benefits relative to the initial installation. The central scenario retains a positive NPV at all three rates. Behavioural sensitivity spans a wider range of outcomes in the cases examined.
5.4 Three practical break-even tests
The required annual benefit is $812 per bus. Solving PV(B) = PV(C) gives B* = C0/A + M. The result is $811.89 under Table 2. This is the annual value of avoided harm needed to cover installation and recurring cost over ten years.
The required baseline trauma burden is $3,653 per bus-year. With the central 22.2 per cent reduction, H* = B*/R. A service with that expected burden reaches break-even when wearing improves from 20 to 60 per cent under the assigned belt effect.
The required achieved wearing rate is 52.5 per cent. Rearranging Equation 1 gives p1* = p0 + B*(1 − p0e)/(He). The calculated increase is 32.5 percentage points above the 20 per cent baseline.
At the central risk and operating assumptions, an installed price of up to $6,321 produces a BCR of at least one. This price ceiling equals the present value of annual benefits after recurring costs: ($1,000 − $100) × 7.0236.
6 Applying the model to Australian fleets
6.1 Populate the model with service-specific evidence
An Australian appraisal needs three linked datasets: passenger exposure and casualties, correct wearing under the compared systems, and equipment costs over time. The casualty dataset supplies expected risk for a defined service population. The behavioural dataset supplies the additional protection. Installation and support records supply the resources needed to obtain it.
| Input | Measurement | Use in the model |
|---|---|---|
| Passenger exposure | Occupied passenger-kilometres by route and service | Match intervention coverage to risk |
| Baseline casualties | Multi-year passenger outcomes, crash type and belt status | Estimate annual expected trauma |
| Correct wearing | Controlled observations linked to validated sensor logs | Estimate p0 and p1 |
| Belt protection | Outcome-specific evidence matched to belt and crash type | Estimate e for each category |
| Lifetime cost | Quotes, installation records, maintenance and replacement schedules | Calculate annual resource flows |
| Performance over time | Faults, bypassing, alert response and correct wearing during service | Estimate annual benefit persistence |
A controlled fleet trial can allocate buses to existing practice, general reminders, sensor-triggered alerts and their combination. The sensor contribution is then estimated within each reminder condition. Analysis accounts for repeated journeys on the same bus and shared operator practices. Comparing the combined system with general reminders answers the incremental purchasing question for services already using those reminders.
For a casualty rate expressed per passenger-kilometre, expected annual casualties equal that rate multiplied by the corresponding passenger-kilometres. This calculation already includes exposure. Baseline casualty counts, wearing rates and intervention coverage must refer to the same passenger population. Service groups with different risks are calculated separately and then summed.
6.2 Resolve the parameters that drive the decision
The wearing sensitivity in Table 4 identifies the most useful behavioural comparison: whether sustained correct use clears the 52.5 per cent threshold under the central assumptions. The $3,653 trauma threshold identifies the necessary risk evidence. Actual installation quotes can be compared directly with the $6,321 ceiling. Each threshold translates an uncertain input into a measurable decision condition.
A full uncertainty analysis varies behavioural change, casualty rates, belt protection, useful life and cost jointly. Each simulation calculates its own annual benefits, BCR and NPV. Reporting the share of simulations with positive NPV shows how frequently the intervention clears break-even under the fitted input distributions. Dependencies between occupancy, route risk and wearing are retained in that analysis.
System design also enters through operating effects. Passenger response, false alarms, driver workload and maintenance demands are measured alongside wearing. Their observed consequences enter the relevant benefit or cost stream. This keeps the economic assessment connected to the way the system operates throughout a journey.
6.3 Move from fleet procurement to national appraisal
A national assessment adds vehicle cohorts over successive years. Each cohort has an installation date, an expected retirement date and a profile of passenger exposure. Existing voluntary installations enter the baseline. Incremental installations receive their own costs and future benefits. This structure allows a requirement for new vehicles and a programme of retrofits to be evaluated on their respective schedules.
The final comparison reports total discounted benefits, total discounted costs, NPV, BCR and the distribution of benefits across service types. Operator expenditure is shown alongside the social result so that purchasing and funding decisions can be assessed together. The economic case follows from the measured improvement, the passengers exposed to risk and the resources required to sustain that improvement.
7 Conclusion
Per-seat seatbelt sensors and alerts can be appraised by connecting additional correct wearing to avoided passenger harm and comparing that benefit with lifetime system cost. The constructed scenario yields a BCR of 1.23, an NPV of $1,321 and an equivalent annual net benefit of $188 per bus. Its break-even conditions are $812 in annual benefits, a baseline trauma burden of $3,653 per bus-year, or achieved correct wearing of 52.5 per cent under the stated assumptions.
These thresholds give Australian fleet research a concrete economic target. Correct-wearing observations, matched casualty exposure and installed-system costs supply the inputs needed to calculate a service-specific result. The same model can then support fleet procurement, comparisons of alert configurations and a national assessment of deployment.
8 References
- Department of Infrastructure and Transport. Regulation Impact Statement for Amendments to Vehicle Standard (Australian Design Rule 4/04 – Seatbelts) 2006: Seatbelt Reminder Systems; Seatbelts for Folding Seats. January 2012. OBPR reference 11243. Sections 5.1 and 5.2, Table 2 and Appendix 3. Original government statement.
- Park J, Lee SH, Jun JH. The Effectiveness of Seat Belt Reminder (SBR)s by Analyzing the Result of the Pilot Project of an Interurban Bus With SBR. Proceedings of the 27th International Technical Conference on the Enhanced Safety of Vehicles. 2023. Paper 23-0190. Methods and results, pp. 2 to 6. Conference paper.
- Office of Impact Analysis. Value of Statistical Life. 4 February 2026. Values expressed in 2025 dollars. Valuation guidance. Accessed 31 July 2026.
- Office of Impact Analysis. Cost Benefit Analysis. Sections on discounting, net present value and sensitivity analysis. Economic appraisal guidance. Accessed 31 July 2026.