How Pure Water Cleans Windows Without Streaking: Inside US Patent 20120325733

Most of what gets called "cleaning automation" is loud: vision stacks that flag dirt, UV stacks that destroy microbes, robots that announce themselves on a sales call. The companion reads on our computer-vision piece and the pulsed-xenon UV piece are both about that kind of thing. This one is about something quieter. Over the last fifteen years the commercial window-cleaning industry walked away from soap. The squeegee is still on a lot of belts, but the actual cleaning chemistry on most commercial glass jobs now is the same chemistry that comes out of a semiconductor fab rinse bath: water with almost nothing dissolved in it.

The cleanest patent capture of the cart-mounted version of that idea is US20120325733A1, Reverse Osmosis and De-Ionized Water Supply for Window Cleaning, filed by Meyer Ostrobrod in June 2011 and published December 2012. The full text is on Google Patents at patents.google.com/patent/US20120325733A1. It's an interesting filing for a reason most patent-read articles wouldn't lead with: it was abandoned in November 2014 for failure to respond to a USPTO office action. The abandonment is the most useful part of the story.

What the Patent Actually Describes

Core System

Strip the legal scaffolding and US20120325733A1 is a wheeled cart with a frame, hose connections, and a stack of filters chained in series:

• Spigot inlet — connects to ordinary building or municipal tap water.
• Sediment filter — removes particulates that would foul the RO membrane.
• Activated charcoal filter — removes chlorine and organics that would degrade the membrane.
• Reverse osmosis unit — semi-permeable membrane that splits the feed water into a permeate stream (clean water that passed through the membrane) and a reject stream (everything that didn't).
• Deionization unit — mixed-bed ion-exchange resin that polishes the RO permeate down to near-zero dissolved solids.
• Output hose to spray nozzle or brush head — for delivery onto the building's glass.

The architecture is what every Tucker, IPC, Streamline, Pure Water Window Cleaning, and Unger HydroPower cart on the market today still looks like. The patent's contribution wasn't inventing those filter stages — it was claiming the combination on a single mobile cart purpose-built for window cleaning.

The One Clever Claim

Reverse osmosis is embarrassingly inefficient. A typical low-pressure RO membrane recovers somewhere between 5% and 15% of its feed water as usable permeate, and rejects the rest. Pump 100 gallons of municipal water into the cart and 85 to 95 gallons leave the membrane as concentrated reject — perfectly clean water you paid for, now with all the building's dissolved minerals stuffed into it. For a window-cleaning operator who pays for water and is also trying to be defensible to a sustainability auditor, that's a problem.

The genuinely non-obvious claim in US20120325733A1 addresses it. The filing describes a two-hose configuration: the reject stream is captured and pumped through a second hose for prewash, then the permeate (polished by DI) goes through the main hose for the final rinse. Mineral content in the reject doesn't matter for the prewash step — you're knocking dirt off the glass, not drying water on it. The expensive purified water is reserved for the rinse, where chemistry actually matters. Same total cart throughput, much less wasted water.

It's the only piece of the filing that isn't standard plumbing.

Abandoned

Status on the Google Patents record reads "Abandoned," with a November 2014 entry recording failure to respond to a USPTO office action. Ostrobrod's professional history sits in industrial fall-protection equipment, not janitorial — this looks like a side filing he didn't follow through on. No assignee picked it up. No continuation was filed.

The result is that the cart-mounted RO+DI architecture is freely commoditized across every major water-fed-pole brand on the market. Nobody owns the category. That fact alone explains more about how this technology spread than any marketing deck does.

Why Pure Water Cleans Glass Without Streaking

This is the part of the story most operators never explain to clients, and it's worth getting right.

Ordinary municipal tap water carries dissolved solids — calcium, magnesium, sodium, silicate, bicarbonate, sulphate — at concentrations of roughly 150 to 400 parts per million in most Canadian cities. When that water dries on glass, the water leaves but the dissolved minerals don't. They precipitate as a fine white haze, which is the "spot" in spot-free. The squeegee evolved as a workaround: drag the water off the glass before it has time to dry and the minerals leave with it.

RO membranes typically remove 95–99% of dissolved solids. A mixed-bed DI unit polishes whatever's left down to under 10 ppm, often under 1 ppm at the meter. Water that clean has almost nothing in it to leave behind. Spray it on glass through a brush, agitate to lift the dirt, rinse, and walk away. The water sheets off and dries clear. No squeegee pass. No towel. No streaks. The chemistry does the work that hand technique used to do.

There's a secondary effect that quietly killed detergent on most commercial glass jobs. Soap residue is itself a streak source — surfactant film left on the glass after the rinse evaporates becomes the next pane's haze. Pure water, by contrast, is mildly aggressive on its own as a solvent because it's chemically "hungry" — water with nothing dissolved in it preferentially dissolves whatever it touches. So pure-water systems clean better and leave nothing behind, which is why the commercial side of the industry quietly stopped buying glass cleaner by the case over the last decade.

The Twist — The Clever Bit Wasn't Even New

The reject-stream-as-prewash claim is the cleverest engineering in Ostrobrod's filing. It also wasn't original. The same idea — RO reject for prewash, permeate for final rinse — was patented for car washes in 1991 by Brite-O-Matic Manufacturing Inc.: US5160430A, Car Wash System Using Reverse Osmosis Concentrate for Initial Rinsing and Permeate for Final Rinsing. Twenty years before Ostrobrod's filing, the same two-stream architecture was already running in commercial car washes for exactly the same reason: spot-free rinse without wasting the reject stream.

So the picture, in operator terms, is this: a fall-protection inventor filed a patent in 2011 that took a 1991 car-wash idea and applied it to building windows on a wheeled cart. The USPTO sent an office action. He didn't respond. The filing went abandoned in 2014. Everyone in window cleaning kept building the same architecture anyway, because nothing in it was strongly defensible to begin with.

That sequence is part of why the industry standardized on cart-mounted RO+DI in the way it did — fast, with multiple brands, with no licensing fees, and with no court fights. Patents shape industries when they're prosecuted and enforced. They also shape industries when they aren't.

Northern Ontario Water Chemistry

The reason any of this matters operationally in our market is that purified-water economics depend on what's in the feed.

Sudbury draws from the Wahnapitae and Ramsey Lake systems, sitting on Canadian Shield bedrock. The water is generally moderate in hardness but carries a measurable mineral and silica load that varies seasonally. North Bay draws from Trout Lake, which is softer on average but still measurable. Either way, every cart full of feed water sends a few hundred parts per million of dissolved solids onto the DI resin, and DI resin has a finite ion-exchange capacity. The harder the local water, the faster the resin saturates, the more often the operator changes media, and the higher the per-gallon cost of finished pure water.

This is the part of water-fed-pole economics that doesn't show up in a sales brochure. The patent describes a clean architecture; the field reality is that operating cost per square foot of glass cleaned tracks directly with feed-water TDS. Sudbury jobs and North Bay jobs run different consumables curves. Anyone quoting commercial window cleaning across both cities without knowing that is guessing.

Where Water-Fed Pole Cleaning Is Actually Going

Three things are visibly moving in the field right now.

Higher-capacity DI resins. The current generation of mixed-bed resin from companies like Purolite and ResinTech is shifting capacity per litre upward, which extends the run time between regens and changes the breakeven point on local water chemistry.

On-board TDS metering with logging. Newer commercial water-fed-pole carts log permeate TDS by minute, so the operator can prove to a client that the water hitting their glass was below a contractual threshold. This is the same audit-trail logic we saw in the pulsed-xenon UV piece — commercial cleaning is being pulled, slowly, toward verifiability.

Carbon-fibre poles past 60 feet. The pole-engineering side of the story has its own patent thread, including GB2565148A on anti-rotation clamp design. Reaching higher means more flex, more clamp wear, and more failure modes the patent literature is now starting to address explicitly.

What hasn't changed, and isn't going to: the part of window cleaning that earns the contract is still surface preparation, edge detail on framed windows, ladder discipline, scaffold work where height demands it, and the boring fact that pure water doesn't dissolve baked-on bird droppings or construction silicone overspray. Those still need a hand.

What Binx Is Actually Doing With This

Binx runs cart-mounted RO+DI water-fed-pole systems for commercial window cleaning across North Bay and Sudbury, and the chemistry described in this patent is the chemistry actually arriving at our clients' glass. The job split is straightforward and we don't oversell either side:

Pure-water WFP does the heavy lifting on multi-storey commercial storefronts, school exteriors, medical and office building facades, post-construction window cleanup, and any exterior glass where ladder time and squeegee time would dominate the labour budget. The cart sits at ground level, the operator works upward with the pole, finished water sheets off, glass dries clear. That's the commercial window cleaning workflow.

Squeegee and hand work still owns the interiors, the framed and divided panes where chemistry alone can't clear the edges, and most of our residential window cleaning routes. Pure water is overkill on the average home window and the cart-setup time doesn't amortize on small jobs. The right tool for the right contract.

Where we apply the patent's actual insight is on the cost side. Tracking permeate TDS against local feed-water TDS, planning resin changes against route geography, and not pretending all glass jobs are the same job — that's the operator's read of US20120325733A1 in practice. We're not trying to claim a technology that the patent literature itself shows is unowned. We're just trying to deliver it competently.

Related Reading

• US20120325733A1 — Reverse Osmosis and De-Ionized Water Supply for Window Cleaning (Meyer Ostrobrod, abandoned 2014 — the primary anchor)
• US5160430A — Car Wash System Using Reverse Osmosis Concentrate for Initial Rinsing and Permeate for Final Rinsing (Brite-O-Matic Manufacturing, 1991 — the foundational prior art that predates the window application by 20 years)
• US7874757B2 — Window Cleaning Apparatus with Deionization Cartridge (Bruggeman — the DI-only architecture that predates cart-mounted RO+DI)
• GB2565148A — Water-Fed Pole for Window Cleaning (UK filing on the anti-rotation clamp problem for telescoping poles)
• US7416361B2 — Extension Pole (cited inside the Ostrobrod filing as prior art)
• How Cleaning Robots Learn to See Dirt: Inside US Patent 20180092499 — the perception side of the cleaning automation story
• Why Hospitals Use Pulsed Xenon UV Robots: Inside US Patent US20170173195A1 — the disinfection side of the same arc

The Operator's Bottom Line

Pure-water window cleaning is one of the few examples in commercial cleaning where chemistry quietly replaced labour without anyone needing to fight about it. The Ostrobrod filing captures the architecture cleanly, the abandonment captures why nobody owns it, and the Brite-O-Matic prior art captures the fact that the cleverest idea in the filing was already running in car washes 20 years earlier. None of that diminishes the technique. It just explains why every serious commercial window cleaner in the country is using essentially the same setup. The patent didn't make the industry. The industry made itself, and the patent literature mostly just describes where it ended up.

---

Binx Professional Cleaning is a commercial cleaning company serving North Bay and Sudbury, Ontario, managing over 500 bathrooms nightly across schools, healthcare facilities, and commercial properties. We operate cart-mounted RO+DI water-fed-pole systems for commercial window cleaning across both cities. Get in touch for a quote.