A friend of mine recently gave birth in the car park of a hospital after going into labour two weeks early. "Didn't you know?" I asked, "that things were happening down there? Didn't you feel any twinges?" After all, this was her second baby. "It was completely different this time," she said. "I had no idea I was in labour until it was too late."

The early stages of labour can certainly be mystifying, whether it's your first baby or your fourth. It is often hard to tell whether the twinges are indigestion, Braxton-Hicks contractions (largely painless "practice contractions" that can happen for weeks running up to the birth) or the start of the real thing.

This age-old uncertainty, however, could soon be at an end. A British company, Jopejo, has developed a device that, they claim, can predict whether labour is about to start. It is simple to use, noninterventionist, and could sell for as little as £20.

The obstetricians who developed the prototype - Dr Nigel Simpson and Professor James Walker of Leeds University's medical school - built up a picture of how electrical activity in the uterus changes during pregnancy. They found these patterns to be common to every woman and discovered that it is therefore possible to distinguish whether the uterus is 48 hours or two weeks away from delivery.

To use Jopejo's device, you place its electrodes on your stomach and a watch the "traffic lights": a red signal means you will not go into labour in the next two weeks; an amber one that you should start contracting within a fortnight and a green one that labour will start within the next 48 hours.

This is more than just a handy way to help anxious mothers to decide whether to rush to hospital or take a Tums. Of approximately 600,000 pregnancies in the UK each year, 10% may be at risk of preterm delivery (before 37 weeks).

Dr Alan Cameron, a consultant obstetrician and foetal medicine specialist at the Queen Mother's hospital, Glasgow, sees the consequences of early-onset labour every day. "Preterm labour is our main problem. In general, once we see it, it's too late. The uterus could have been contracting for days, even weeks, beforehand."

These so-called "silent labours" can be quite advanced by the time the woman gets to hospital and often there is nothing that can be done to prevent the baby from being born. A machine, therefore, that could be used on women deemed to be at risk of premature delivery could have a massive impact on foetal survival rates. "A reliable device like this," says Dr Cameron, "would mean we could hospitalise the mother, treat any infection and give drugs that may prevent the contractions from starting."

Knowing that labour will begin in the next two weeks (or even 48 hours) would also give doctors valuable time to treat the woman with steroids and thereby help her premature baby's lungs to develop.

Melanie Every of the Royal College of Midwives is, however, "sceptical about how accurate such a device could possibly be". In the case of normal pregnancies, any degree of inaccuracy, however small "could cause women considerable distress and disappointment". Apart from the emotional consequences of raising false expectations, there is also, she says "some evidence relating stress to the progress of labour".

Lucy Shaw is two weeks away from her first baby's "due date". Would she buy the machine? "I rather like the uncertainty of it all," she says. "It's a bit like not knowing the sex: there's something fun about the waiting." She concedes, however, that, "If I was a week late and facing induction, I'm sure I'd buy it just to rule out the worry."

Certainly my friend, in her freezing car park, could have done with a warning. But it is more important that the device may have the capacity to save a baby's life. If it does, it will truly be a landmark in antenatal care.