The Monorail to Nowhere: What the Kurtsystem Revealed About Automating Horsemanship
- Aug 24
- 4 min read
A Turkish owner spent tens of millions building a monorail system to exercise young racehorses without riders, betting that removing human error from early training would prevent injury. Nine years after it opened at Kingwood Stud, the stud is for sale, and the system built to replace horsemanship is now a line item in the demolition quote.
By Aleksandra Spasic
Published August 24, 2026
Kingwood Stud is for sale. Tucked into the agent's particulars for the 158-acre Lambourn property, past the boxes and the paddocks and the six-furlong covered gallop, is a small, clinical line about the estate's most expensive feature: the pods designed to help rehab and control the horses are no longer in use. The agents have obtained quotes for tearing the whole structure down, or at minimum stripping out its internal rails. Those quotes sit in a data room, waiting for a buyer to decide how much it costs to remove a machine that was supposed to change how racehorses are made.
That machine is the Kurt system, and its arc from unveiling to teardown quote is worth examining closely. The real story here isn't a failed gadget. It's what happens when a legitimate welfare concern gets answered with an engineering solution instead of a horsemanship one.
The premise was never in question. Two-year-old thoroughbreds carry underdeveloped musculoskeletal systems into hard training, and the resulting tendon and ligament injuries are a well-documented source of wastage in the sport. Turkish industrialist Mehmet Kurt, who had bred and raced horses to two Turkish Derby wins, spent roughly two decades developing an answer: an automated pre-training system that would exercise young horses without a rider on their back at all. A prototype ran on his farm in southern Turkey, reportedly cycling more than a thousand horses through it. In 2011 he bought Kingwood Stud, a Lambourn property with real pedigree of its own, dating to 1924 and the breeding of a Derby winner in Felstead. By early 2017 the finished version was open: a covered monorail running an oval sand and wax canter of just under a mile, with up to ten horses moving simultaneously in individual cabins, tethered rather than ridden, wired to equipment tracking heart rate and gait, along with respiration. Estimates of the total spend range from roughly £20 million in UK coverage to as high as $46 million in the American trade press. Even the local MP called the investment a vote of confidence in the village.
What the system was actually built to remove tells you where the design logic went wrong. A traditional hot walker, the mechanical arm-and-lead-rope device found in yards worldwide, does one narrow job cheaply: it moves a horse in a circle at a fixed pace so it can cool down or get low-grade exercise without tying up a member of staff. Nobody markets it as an alternative to riding, because it was never meant to be one. The Kurt system was different. Its stated purpose was to substitute for the early, formative stage of ridden work itself, the stage where a green two-year-old first learns to carry weight and respond, in real time, to a rider's balance and judgment. That is not a mechanical problem. Starting a young horse is a negotiation conducted through feel and constant small correction. That is precisely the kind of nuanced, responsive work a fixed-speed monorail cabin cannot perform, no matter how sophisticated its biometric readout. A machine can move a horse forward, but it cannot read the hesitation in a stride or soften its hand when a horse gets anxious. It cannot make the split-second judgment call that separates a good young horse's education from a bad one.
The skepticism was there from the beginning, if quieter than the trade coverage suggested. Public reaction when the system was unveiled ranged from confusion over why the industry's standard methods needed replacing at all, to specific welfare worries about how ten tethered horses in a fixed-speed rail system would be safely stopped in an emergency. One trainer's observation, offered dryly in the comments of an early write-up, cut to the point better than any press release did: a longe line achieves much the same result for a fraction of the cost, using a human being who can actually see and respond to the horse. Kurt Systems' own spokesperson leaned on the claim that the design removed the possibility of jockey error, as though rider judgment were purely a liability to be engineered out rather than the mechanism by which a young horse is actually taught anything. The company maintained that no horse trained in the system had ever fallen, and that horses emerged stronger and better balanced. It never published comparative injury data to support that claim, and no independent study appears to have tested it. A welfare case built entirely on the word of the company selling the product is not a welfare case at all.
Nine years on, the outcome is written into the sale listing itself. The rehabilitation pods sit unused. The structure carries a rateable business tax value of £53,000 a year, money the estate's next owner would rather not keep paying for equipment nobody is running. Removing it is framed as a straightforward cost saving, not a loss.
None of this means racing should give up on automation. The Kurt system's fate doesn't prove technology has no place in reducing injury among young horses. Biometric monitoring and wearable sensor technology have found real, lasting adoption elsewhere in the sport precisely because they were built to support a rider's judgment rather than replace it. Racing's actual welfare gains over the past decade have come from giving experienced horsemen and horsewomen better information, not from taking the horseman out of the loop. Kurt's monorail tried to automate the one part of early training that most depends on a human being able to feel what a horse needs and respond to it in real time. It could raise a horse's heart rate and record its stride length. It could not replace the rider whose hands were shaping how that horse would carry a rider for the rest of its career.
A stud with a hundred years of breeding history is now being sold with a demolition quote attached to its most expensive addition. What should the industry be willing to spend to solve a welfare problem it already understands how to solve, and what does it cost when the money goes toward removing the horseman instead of supporting one?