Canadian AI Cuts Solar Panel Cleaning Costs by Millions

Solar farms lose millions every year because they clean their panels either too often or not often enough, but a Canadian startup just cracked the code on perfect timing.

Swish Solar launched SwishOS, an artificial intelligence platform that tells solar farm operators the exact moment when cleaning their panels will pay off. The software calculates when dust buildup costs more in lost energy than a cleaning crew would cost to fix it.

"Operations teams typically rely on a fixed schedule for cleaning, regardless of the degree of soiling," says co-founder Miswar Syed, who grew up watching solar farms struggle with this problem in Saudi Arabia. "The costs of robots, transportation, distilled water, and field workers add up."

The platform works without any new equipment. It analyzes existing solar farm data along with local costs for water, electricity, and labor to build custom cleaning schedules that adapt to changing weather and dust conditions.

Instead of cleaning panels every Tuesday whether they need it or not, operators can now see a running total of how much money dirty panels are costing them. When that number exceeds the cleaning cost, the software sends an alert.

The technology has already attracted customers with solar assets across North America and the Middle East. Founded in 2024 by Syed and Amirhossein Boreiri, the company raised $1.5 million in pre-seed funding and has four patents pending.

The Ripple Effect

Better cleaning schedules mean solar farms generate more electricity without wasting water or labor on unnecessary cleanings. In water-scarce regions like the Middle East, avoiding even one unnecessary cleaning across a large solar farm can save thousands of gallons of distilled water.

The financial impact scales quickly. A single large solar farm that optimizes its cleaning schedule could recover hundreds of thousands in lost energy production while cutting maintenance costs.

Swish Solar is also developing SwishScreen, a polymer film that uses electric charges to shake off dust and snow without any water or mechanical cleaning. Early prototypes are already being tested in Saudi Arabian desert conditions, where dust storms regularly coat solar panels in thick layers of grime.

The company prices its software on a sliding scale based on plant size, with discounted rates for smaller systems under one megawatt. This approach helps community solar projects and smaller installations access the same optimization tools as massive utility-scale farms.

As solar energy expands into dustier climates where it's needed most, smart cleaning could be the difference between profitable clean energy and panels too dirty to pay for themselves.