AssetID: 54911393
Headline: RAW VIDEO: AI-Powered Underwater Jet Pack Allows Divers To Swim Like James Bond
Caption: A Hong Kong-based start-up has launched a crowdfunding campaign to create an affordable James Bond-style underwater jetpack - rather like 007 used in the film Thunderball. XiaoTun’s AJ-03, also referred to as the Smart Diving Jetpack or AquaticJetpack is not the first submersible jetpack to hit the market, but it introduces several standout features that differentiate it from its competitors. Most notably, while many underwater propulsion systems are worn on the arms or hips, XiaoTun’s model is worn like a backpack — similar in style to the $18,000 CudaJet, but offered at a fraction of the cost. Weighing just over 9kg (20lbs), the AJ-03 is powered by a pair of jet modules capable of generating up to 100Nm of torque each. It runs on a 15,000mAh lithium iron phosphate battery, offering between 30 and 90 minutes of use depending on the selected speed. The jetpack is secured to the diver with a chest harness, complete with shoulder and neck straps. There is also room to attach a two-litre scuba tank, although compatibility with additional dive gear such as buoyancy control devices or weight belts remains unclear. The jetpack is rated for use at depths of up to 20 metres (66 feet) and features a tethered remote control with a built-in battery indicator, allowing divers to move forwards, backwards, or rotate with the press of a button. For hands-free operation, users can engage an AI-powered cruise control mode that uses body movement to determine direction. Three pre-set speed settings are available: 0.59m/s, 1m/s, and 1.5m/s. XiaoTun’s Kickstarter campaign runs until Friday, 25 July. Early supporters can secure the AJ-03 for $1,374 (£1,060), almost half the projected retail price of $2,727 (£2,100). If successful, the AJ-03 could be a game-changer for recreational divers seeking a more intuitive and streamlined underwater experience.
Keywords: XiaoTun,jetpack,feature,photo,video,underwater,diving
PersonInImage: