The Center for Data Innovation recently spoke with Ari Isaak, the founder of Photometric AI, a San Diego-based company developing an AI-powered platform that designs and manages street-lighting systems. Isaak explained how the platform combines mapping data, lighting models, and real‑time information to optimize how individual streetlights operate, helping cities improve safety and lower energy use.
David Kertai: What problem is Photometrics AI solving?
Ari Isaak: Streetlights frequently malfunction, such as turning on or off at the wrong times, spilling light into homes, or flooding fields and natural areas, even though detailed lighting specifications and remote controls exist. The industry relies on a handful of cookie‑cutter designs applied across tens of thousands of fixtures, and these templates ignore real‑world variation such as curves, intersections, bike lanes, sidewalks, and elevation changes. A design meant for a straight roadway can create glare or dark spots when installed near an intersection or curve. Even small mismatches reduce visibility, waste energy, contribute to accidents, and create unnecessary light pollution.
Photometrics AI addresses this by designing every streetlight individually at network scale, using mapping technology and AI models delivered through our online platform. The platform uses manufacturer‑provided photometric files, which describe how each fixture spreads light, along with detailed location data to ensure the right amount of light reaches roads and sidewalks. By tailoring each fixture’s performance to its exact surroundings, the system delivers safer, more efficient lighting.
Kertai: What data does your system use?
Isaak: We use geographic information system (GIS) data, including road widths, sidewalks, pole locations, mounting heights, aerial imagery, tree inventories, and crash or crime histories, and combine it with manufacturer‑provided lighting files that describe how each fixture distributes light.
The platform also integrates live data so lighting can adapt to changing conditions. It pulls in bird‑migration forecasts from systems like BirdCast, enabling cities to dim lights only in specific areas during active migration periods. Weather, traffic, and event data further guide adjustments for fog, snow, concerts, sporting events, or emergency‑response situations. Together, these inputs allow the system to determine the safest and most efficient lighting configuration at any moment.
Kertai: How do your AI models optimize that data for lighting urban areas?
Isaak: Photometrics AI optimizes city lighting by starting with the lighting levels recommended in national roadway‑safety standards, such as IES RP‑8, and determining how each streetlight should operate to meet those targets. A busy intersection may require brighter lighting during evening traffic, while the same location can safely run at lower brightness at 3 a.m. when activity drops. The system continuously recalculates lighting as conditions change due to rain, snow, traffic patterns, or large public events.
To perform these calculations across an entire city, the platform uses AI pattern‑recognition models to accelerate the process. Instead of recalculating every light from scratch, the models identify recurring lighting patterns and apply them quickly while maintaining near‑computational accuracy. This approach allows the system to design about seven lights per second, enabling cities to update thousands of lights fast enough to respond to real‑world conditions.
Kertai: How do cities and utilities use the insights to make decisions?
Isaak: The platform helps cities optimize lighting performance and improve long‑term planning. Many cities run streetlights at full power continuously to reduce liability concerns, even though that approach creates glare, wastes energy, and produces uneven lighting. Because Photometrics AI verifies compliance with roadway‑lighting safety standards for every individual light and street segment, operators can confidently adjust brightness levels without compromising safety.
This approach reduces glare, lowers maintenance costs, extends fixture life, and decreases energy use. Instead of relying on assumptions, cities can see exactly how their lighting systems perform and make evidence‑based adjustments through the platform.
The data also supports broader planning goals. Officials can examine lighting performance in areas with high crash rates, speeding concerns, economic‑development projects, or underserved neighborhoods. The system evaluates the entire network consistently; it reveals lighting inequities between communities and identifies areas that need infrastructure improvements. Policymakers gain concrete data they can use to improve safety, visibility, and quality of life more fairly across the city.
Kertai: What gaps has your work revealed?
Isaak: Street lighting affects transportation safety, public safety, energy use, conservation, and grid operations, yet cities and utilities often manage it as static infrastructure. Most U.S. cities still pay fixed electricity rates for streetlights regardless of how much energy they actually save. When a city reduces lighting energy use through optimization, the financial benefits often don’t return to the city.
Our work shows how incomplete that framework is. By measuring performance for every individual light and roadway section, we can quantify benefits cities often overlook: reduced glare, fewer maintenance trips, longer fixture life, improved safety, environmental protection, and the ability to reduce electricity demand during grid stress. In many cases, these operational and public‑safety benefits create more value than the direct energy savings.
That’s why our work increasingly includes policy discussions alongside technology development. We help cities and regulators rethink how streetlights are measured, managed, and valued so communities can capture the full benefits of smarter lighting systems.
