Stage 1
Diagnosis
We establish that the municipality is lighting sub-optimally. In our experience it always is. The finding is delivered in a 60-page report with a heat map of streets lit beyond their economic justification.
A street light costs money every hour it burns. A dark street costs money only when something happens on it. Until now, municipalities have resolved this with a timer and a feeling. LUX-ROI replaces the feeling.
The model joins the real-time price of electricity with the local value of a statistical life, weighted by district income, the value of the vehicles parked on the street and the probability that a victim will file a complaint. Each street receives an individual switch-off hour, calculated to the minute. The precision, we stress, is a property of the model, not of the data.
LUX-ROI does not make political decisions. Political decisions are made by the parameters.
Method. Hourly cost of light = lamps × lamp power × electricity price. Hourly expected harm = lamps × base fatal-incident rate per lamp-hour × hourly activity profile × share attributed to darkness × value of a statistical life × district income weight × Premium Citizen coefficient. The light stays on for every hour in which expected harm exceeds its cost. The base rate is national road deaths at night divided by national lamp-hours of darkness. The activity profile is a fixed curve (3.0 at 17-19 h, 2.0 at 19-21 h, 1.0 at 21-23 h, 0.5 at 23-01 h, 0.2 at 01-05 h, 0.8 at 05-07 h) chosen by Mandate Advisory. Clients may purchase a different curve.
| Life valued at | Darkness blamed for | Switch-off |
|---|---|---|
| $13.7 million (United States) | 50% of night incidents | Never. The light pays all night. |
| $13.7 million (United States) | 7% (clock-change study) | 01:00, back on at 05:00 |
| $0.67 million (Poland, 2.6 million PLN) | 50% | 21:00 |
| $0.67 million, district income weight 0.5 | 50% | 19:00 |
| Any | 0% (England and Wales study) | Sunset |
Forty lamps, 70 W, $0.25 per kWh. The hour is not a property of the street. It is a property of whose life, and which study, the council selected.
We establish that the municipality is lighting sub-optimally. In our experience it always is. The finding is delivered in a 60-page report with a heat map of streets lit beyond their economic justification.
We state, street by street, when to switch off. The schedule is uploaded to the lighting controller and to Annex 7 of the budget resolution, where it becomes publicly available and therefore accepted by residents.
For a separate fee we verify that the municipality switched off as recommended. Clients who omit the audit forfeit the outcome guarantee. The guarantee covers the model's outcome, not the street's.
We understand that switching off a street light on the basis of an equation may raise questions from families, prosecutors and local journalists. The municipality is not left alone with the model.
Every subscription includes complete documentation establishing that the decision to switch off was objective, data-driven and taken by an algorithm rather than by anyone in particular. Every incident after dark receives a Certificate of Model Compliance confirming that it occurred during an hour in which it was economically justified.
The package includes litigation. In court our partner counsel show that the injured party, by walking down an unlit street, accepted the risk coefficient assigned to their district, which was publicly available in Annex 7 of the budget resolution.
Where a case nonetheless ends in damages, the award is entered into the model as the cost of the incident. This raises the value of life in that district and extends its lighting the following year. We call this learning by verdict.
Legal Shield does not cover claims arising from model error, because the model does not make errors. Errors are made by reality, when it departs from the assumptions.