Windshield Wiper

Mount a sheet of glass or acrylic in a frame on the bench. Tape the glass edges. Never develop this on a car you intend to drive.

Decide the pivot point at one lower corner and scribe the arc the blade will follow. Everything else is sized from this arc.

Drill the frame at the pivot and fit a close-running shaft so it turns freely without side play. Slop here shows up as chatter across the whole sweep.

Cut a flat steel arm long enough to reach across the pane, with a slight bend so it stands off the glass and carries the blade parallel to it.

Form a channel that grips a rubber strip along its length and pivots on the arm. The blade must be free to rock a few degrees, or it cannot follow curvature.

Cut a strip of rubber and trim one long edge dead straight with a sharp knife. The edge does all the work — a ragged edge leaves a smear on every pass.

Add a spring pulling the arm toward the pane. Enough force to keep contact across the whole arc, no more — excess pressure judders and tears the rubber.

Fix a lever to the same shaft on the inner face of the frame, so turning it inside swings the arm outside. This through-shaft is the heart of Anderson's claim.

Fit a washer or gasket where the shaft passes through. A hole in a windscreen frame that lets water in has solved nothing.

Fit a weight on the lever's opposite side so the arm returns to its rest position when released. Anderson's patent uses a counterweight rather than a return spring.

Adjust the holder so the rubber meets the pane slightly tilted, trailing the direction of travel. A blade held square chatters instead of wiping.

Wet the glass before every test sweep. Dry wiping grabs and skips, and it destroys the rubber edge in a handful of passes.

Sweep slowly and watch for gaps where the rubber lifts. Lift at one end means the arm is not tracking the curvature — adjust the bend, not the spring.

Run it repeatedly and inspect the rubber edge. Rounding or tearing at one point tells you exactly where the pressure is uneven.

The patent. Mary Anderson of Birmingham, Alabama filed on 18 June 1903 and was granted US Patent 743,801 on 10 November 1903 for a "window-cleaning device": a lever inside the vehicle operating a spring-loaded, rubber-edged arm outside, returned by a counterweight. She built a working model and had it manufactured. Her patent ran its seventeen-year term and expired in 1920, just as the motor industry was fitting wipers as standard, and she earned nothing from it.

It was rejected as a distraction. A Canadian firm turned it down on the grounds that a moving arm in the driver's field of view would be more dangerous than the weather. That objection is not silly — it is the same argument later made about every dashboard display — but it was wrong, and the reason it was wrong is that a driver's eyes accommodate to a regular, predictable motion far better than to an obscured windscreen.

Why a flexible blade beats a rigid one. Glass in a windscreen is curved, and a rigid scraper can only touch a curved surface along a line at one position of the arc. A rubber strip in a pivoting holder deforms to the local curvature and maintains a continuous contact line all the way across, so it sweeps water ahead of itself rather than under itself. The blade must also trail — held at a slight angle to the direction of travel — so it flexes over and wipes rather than digging in and juddering.

What came after. Anderson's device is hand-operated. Electric wipers followed in the 1920s, and the intermittent wiper was patented by Robert Kearns in the 1960s — who then spent decades litigating against Ford and Chrysler, and won. Wipers are, remarkably, the subject of two of the most famous cases of an independent inventor being denied the value of a patent. The mechanism is trivial; the history is not.