Solar Unsoiled

Solar Unsoiled is a Durham, North Carolina-based software company addressing soiling — the accumulation of dust, pollen and other particles on photovoltaic panels, which reduces energy output and therefore revenue for solar operators. The platform combines four components: a proprietary microscopy method in which technicians use a handheld microscope with a digital camera to measure soiling on site (positioned as a cheaper and faster alternative to IV curve tracing); operational analytics that use machine learning to isolate soiling losses from a site's performance data; predictive modeling that blends historical performance with local climate and atmospheric data from a NASA global climate model to build site-specific soiling forecasts; and proactive optimization that factors in power purchase agreements and cleaning costs to recommend cleaning schedules across a fleet.

The company states that its optimized cleaning schedules can raise annual profit by $1,000–$4,000 per MW and that it works across more than 15 GW of solar assets. It does not perform panel cleaning itself, but connects sites to cleaning providers, including through a cleaning vendor marketplace for sourcing and comparing quotes. Additional product features include a Proactive Planner tool for prioritization, forecasting and budgeting of cleanings, and a Rain Risk feature that uses probabilistic rain predictions supplied by Salient Predictions. Published customer commentary comes from Cypress Creek Renewables and Cherry Street Energy.

Founding story

The company originated from research at Duke University. CEO Michael Valerino, who holds an undergraduate chemical engineering degree, first encountered the soiling problem during a 2016 graduate school interview with Duke professor Mike Bergin, who had observed cloudy solar panels at an air pollution research site in India. Valerino focused his environmental engineering PhD on the science of soiling with the aim of translating it into industry applications, and worked with solar industry practitioners while finishing his doctorate to validate that cleaning decisions were a common operational pain point. Co-founders include CTO Alexandra Rivera and scientific advisors Mike Bergin and Drew Shindell, both Duke researchers.

Business model

Solar Unsoiled provides a software platform to solar asset owners and service providers rather than performing panel cleaning itself; it also routes customers to a network of cleaning vendors through a marketplace for quotes.

Traction

The company states it works across more than 15 GW of solar assets and reports customer-attributed benefits including fleet-wide prioritization of washes and budgeting of crews up to three quarters in advance. Its microscopy method has been developed from approximately 40,000 images. A third-party business database lists approximately 8 employees as of August 2026.

Latest developments

Recent product work described by the company includes the Proactive Planner tool for fleet-wide cleaning prioritization, forecasting and budgeting; a cleaning vendor marketplace for sourcing and comparing quotes; a Rain Risk feature powered by Salient Predictions; and research collaboration with the National Renewable Energy Laboratory and university partners on pollen-driven soiling losses in the Southeast.