The Lacing Engine

A black boot with loose laces above, and the same boot drawn tight by a hand-held stud below
Loose, and then not. The stud on its lanyard is the whole of the controls; the curled hooks up the facing are Mark VI lace guides, which replaced closed eyes so that a jammed lace could be cleared without unthreading the boot.

Northampton had been making boots for three centuries before it was asked to consider a boot that laced itself. On the afternoon of 14 December 1932, in the fitting room of a small works at Far Cotton, Archibald Tinker set a size nine on a stool, pressed a brass stud no bigger than a farthing, and stood back. Four feet of waxed cotton drew itself through fourteen eyelets and finished in a bow. Eleven seconds. Three of the seven men present were timing it; the third had forgotten to start his watch, and asked for the whole thing again.

The device was called the Automatic Shoe Lacer, which is the plainest name Tinker ever gave anything he made. It was a brass housing three inches by one and a quarter, screwed through the outside quarter of the boot, holding a clock spring, a worm gear and a two-lobed cam. The spring was wound by hand each morning, the way a man winds a watch. Everything after that the boot did on its own.

The stud sat on a short lanyard at the end of a sheathed wire, so that nobody had to bend down — the arrangement Ernest Bowden patented in 1896 for bicycle brakes and which remains the correct answer to the problem of pulling on something round a corner. A Bowden cable costs nothing, weighs nothing and cannot be improved on, and Tinker fitted one to every machine he built, first to last. Tension came off the spring at eleven pounds and could be let down to seven for a swollen foot. The laces themselves were ordinary, bought by the gross from a supplier in Kettering. Tinker specified nothing about them except the aglets, on which he would not be moved. He said so in every interview he gave, which is to say once.

Forty-Seven Prototypes

Studio photograph of a man's legs in two black wingtip shoes joined by a rigid cable across the ankles
The Mark IV, photographed for the prospectus. One engine served both boots down the rigid cable across the ankles. The white socks were the photographer's idea and Tinker objected to them in writing.

Nothing about the eleven seconds arrived quickly. Between the spring of 1929 and the demonstration of 1932, Tinker built forty-seven machines and destroyed most of them. The Mark I weighed two pounds six ounces on the foot, which is a little more than the foot: a fractional-horsepower motor and a wet accumulator strapped above the instep, wired to a boot that it laced perfectly and then made unwearable. The trade papers had been invited. One of them printed the word preposterous twice on the same page, and it was within its rights: Heath Robinson had been in use as a common noun since 1917 — that is the earliest citation the Oxford English Dictionary carries for it — for precisely this class of object, and the Mark I earned it honestly.

A pair of black shoes standing on a solicitor's desk beside papers, with a third shoe at the near edge and the agreement being signed at the far end
Gold Street, October 1932. The shoe on the near edge belonged to the clerk and formed no part of the agreement; the pair on the desk was never worn again.

Letters patent came through that autumn. A fortnight later the licence was drafted in a solicitor’s office in Gold Street, samples stood up at one end of the table and the terms argued at the other. That is very nearly the whole commercial history of the device before the war — it was admired at close range by men who never put it on.

The knot was the last thing to come right, and it cost three years. A cam cut symmetrically, which is the natural way to cut a cam and the way any competent shop would cut it, draws the second half of the stroke in the same direction as the first. That ties a granny. A granny stands up along the shoe instead of lying across it and walks itself loose inside a mile, and it does so for reasons that were not measured until 2017, when a Berkeley group put accelerometers on the knot and found six to eight g going through it at every heel strike while the swinging free ends pulled the centre open. Tinker had no accelerometers. He had complaints, and for eighteen months he could not tell his own fault from the wearer’s, since a man who has tied a granny since he was five will tie one over the top of anything you sell him. The cure was to reverse the second lobe of the cam. He recut it in an afternoon.

The Bureau Calls at the Works

Two men in tweed seated facing each other in a parlour with small brass mechanisms on the carpet between their boots
Tinker and the Bureau's correspondent, with two engines set out on the carpet between them. The men at the window are from the works; the one with the notebook is not.

Tinker gave one interview in his life, in the fitting room at Far Cotton on the afternoon of the demonstration, with a coal fire going and eleven pairs of boots on the bench. Our correspondent had been promised twenty minutes and was given ninety. The transcript below is cut only where the fire was seen to.

Portrait of Archibald Tinker

Archibald Tinker

Inventor · Far Cotton, Northampton

Fitting room, 14 December 1932. Two Mark VII units on the carpet, wound and not yet fired.

The Bureau

Everyone begins with the nephew. Begin with the nephew.

Tinker

Wilfred. Six years old, thirty-five minutes on a pair of boots at the foot of the stairs while the rest of the house stood about in the hall in their coats. I timed it twice, in case I was being unfair to the boy. A man may watch that once. Twice and he goes to his bench.

The Bureau

And what came off the bench?

Tinker

Clock spring, worm gear, two-lobed cam, the whole of it in brass and not one part of it new. You wind it in the morning with the heel of your hand, the way you wind a watch, and then you press the stud and stand still. It laces the boot and it ties the bow. It does not polish the boot. I have never once said it polishes the boot.

The Bureau

You are said to be particular about laces.

Tinker

I am particular about one thing in a lace and indifferent to the rest. Hasketh at Kettering has sent me waxed cotton by the gross since 1930 and I have never once specified a colour. I specify the tag. Brass, drawn and rolled, not pressed tin. Tin burrs inside a fortnight and stops in the guide at the third eyelet, and then a man is standing in his own hall at eight in the morning with a boot he cannot get into and a machine he paid four guineas for. Hasketh sent a tin batch through in the spring to see whether I would notice. I noticed on the Tuesday.

The Bureau

The Mark Four is the one in the photographs, with the cable across the ankles.

Tinker

One engine, two boots, and a cable between them to save the cost of a second engine. It worked exactly as drawn. It also fixed the wearer's stride at nine inches, which is a stride for a man crossing a room and no use whatever to a man crossing a road. We sold four pairs. I bought three of them back and I paid the full price, because they were not faulty. They were finished, and wrong.

The Bureau

The licence was signed in October. What have the gentlemen done with it?

Tinker

One of them has a lady's button boot with thirty-one buttons and no lace anywhere upon it, and would like to know what I intend to do about that. One wants it on the strap of a portmanteau. The third has asked twice whether it can be made to undo the bow instead of tie it, which is the only intelligent question anybody has put to me this year, and the answer is that it can and that I will not sell it, because a boot that unties itself will untie itself in the street. Between the three of them they have not yet made a single pair.

The Bureau

A plain question, then. Was any of it necessary? A man in good health ties his own boot in twenty seconds.

Tinker

You may ask me that and I shall not answer it. I have watched three men this year put that question to a woman who cannot reach her own foot, and not one of them stayed for the reply. Ask me the other one. Ask me whether it works, and I shall do it here on the hearthrug in front of you, and you may hold the watch yourself.

The Bureau

Advice, then, for the aspiring inventor.

Tinker

Pick a nuisance small enough that you can stand and watch it happen in front of you. Then be ready to be laughed at in a trade paper for four years. I would do all of it again. I would not do the Mark Four again. And you may finish on this, which I have been saving since August. The future is untied!

Our correspondent left with a Mark VII in his coat pocket and lost it on the train at Blisworth. Tinker posted a replacement the following Tuesday with a note observing that it was the eighth he had sent out that year, and that he did not consider this a good advertisement for the fastening.

A Hundred Years On

A wood-cased maker's plate with a chrome dome and a red button, reading ARCCHIALED TINKER, and a hand lowering a small chrome mechanism onto a brown canvas high-top with a red pull-loop
The Heritage pattern going down onto a canvas high-top, and the badge that comes with it. TinkerTech reproduces the maker's plate exactly as the Far Cotton foundry misstruck it in 1932 — ARCCHIALED TINKER, error and all — and has twice had to explain in writing that this is deliberate.

TinkerTech marked the centenary of the Far Cotton demonstration by doing the whole thing over. The LaceGenius 2.0, announced for 2032, keeps the arrangement Tinker settled in a single afternoon — a stud, a sheathed line, a cam that reverses on the second stroke — and replaces every component of it. The housing is a sealed disc on the quarter. The cam is a micro-actuator the width of a shilling.

What the shilling buys is the one thing a clock spring could never do. The tension is not set by the wearer at all: it is read off the foot forty times a second and trimmed while the man is still walking, harder on the climb and slacker on the flat, because a clock spring does not know it is going uphill.

Two elderly hands pressing a small round control unit clipped to the laces of a grey trainer on a table
Two hands on a button that needs one. The trial found the second hand was steadying the shoe rather than pressing, and the stud moved twenty millimetres outboard for the following case.

The company does not sell it on speed. Its own trial, run over eleven weeks with three hundred and forty participants, found that the people who valued the thing most were not in any hurry: ninety-one of them could not tie a lace unaided at all — arthritis, tremor, a prosthesis, a back that would not fold — and for that group the device is not a convenience but the difference between dressing yourself and waiting to be dressed. Nike’s HyperAdapt 1.0 went on sale on 28 November 2016 as an expensive curiosity for people who did not need it, and its descendants are now sold, without any ceremony whatever, to people who do.

The Works Rebuilt

Dusk exterior of a long low glazed factory with ivy on the piers, roof solar and three vertical-axis turbines above the ridge
The works from the access road at dusk, the hall lit through the glazing. The reeds in front are a balancing pond: the ground at Far Cotton floods, and flooded in Tinker's day as well.

Far Cotton kept the ground and lost the shed. The hall that turns out the LaceGenius 2.0 stands on the site of the fitting room, and is bolted together out of cross-laminated timber rather than framed in steel because the company means to take it apart again: every beam is numbered for a second building that nobody has yet designed. The vertical-axis machines over the ridge were chosen over the ordinary sort because they need not be turned into the wind and can be serviced without anybody leaving the roof. That the whole line should be legible from a public road at all is either candour or advertising, and is probably both.

Inside, the arms do the threading and the men on the walkway are not watching the arms. They are watching one figure: the tension in the cord at the instant the bow closes, cell by cell, which is the only number in the building that decides whether a shoe is finished or scrap. They have stopped the line nine times in eleven months, and seven of those were the same gripper.

Macro view of a small brushed steel disc actuator held in precision tooling with two spools of white filament feeding into it
The unit in the jig. Cord is drawn off both spools at once so that the stroke comes up balanced — Tinker did the same with one spring and a great deal more noise.

The actuator itself is put together at the far end of the line, in a cell no larger than a bread oven: rather less room than Tinker had at his own hearth, doing rather more than he could do there with a file. The finished unit is sealed for the life of the shoe and is never opened again. Tinker’s brass housing would not have taken it as a component. It would have taken it as a decoration.

Whether any of it was ever necessary is the question Tinker was asked in 1932 and declined to answer. He preferred the other one — whether it works — and on that the record is straightforward. A boot at Far Cotton is still laced in about eleven seconds, and on the second stroke it is still turned the other way.