China's new tech for quieter subs could revolutionise underwater stealth: study
Zhang Zhiyi and his team of researchers, from Shanghai Jiao Tong University's State Key Laboratory of Mechanical Systems and Vibration, said their innovation could theoretically cut the detection range of hostile sonar systems by more than half.
The technology is intended to replace traditional rigid engine mounts with a hybrid active-passive vibration isolation system that would reduce the engine noise transmitted through submarine hulls by up to 26 decibels (dB), according to the paper.
It combined a steel-rubber-steel 'sandwich' ring to dissipate vibrations, and an active layer with 12 piezoelectric actuators arranged radially around the engine to further reduce the noise, it said.
These electrically-powered actuators can neutralise the engine's micron-level movements with strong control forces through precision lever mechanisms.
'Research indicates that a mere 10dB reduction in underwater vehicle noise can decrease its detectable range by 32 per cent,' according to the paper, published in April.
'During low-speed navigation, mechanical noise generated by power equipment operation constitutes the primary noise source for underwater vehicles, serving as their key acoustic signature for detection – typically manifested as a series of low-frequency tonal components,' the researchers said.
'Implementing vibration damping measures to diminish energy transmission from engines through supporting structures is crucial for enhancing underwater vehicles' acoustic stealth performance.'
Zhang and his team conducted laboratory tests on a scaled-down prototype that showed unprecedented performance with 24dB reduction (12dB passive plus 12dB active) at 100 Hz, and 26dB at 400Hz.
Effective bandwidth went across 10–500Hz, covering most typical engine harmonics, while real-time noise cancellation was possible thanks to the system's extremely short response time, according to the study.
The researchers said the system's smart FX-LMS adaptive algorithm used a multidimensional control matrix to coordinate all 12 actuators, avoiding destabilising feedback loops – a major challenge in real life applications.
While the lab results are striking, hurdles remain. Rubber stiffness, for instance, shifts under changing temperature and pressures, and it is unknown how durable the piezoelectric materials will prove to be in long-term operations.
It is unclear what type of engine noise reduction is used in Chinese submarines, and the the researchers did not specify the overall improvement on a sub's noise profile that would be achieved by the new technology.
China State Shipbuilding Corporation, a major submarine builder, contributed to the study. - South China Morning Post
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
The Star
8 hours ago
- The Star
Trump administration issues plan to limit AI exports to China
The White House on Wednesday released proposals to restrict exports of American AI equipment and limit the spread of Chinese AI models as part of a sweeping plan to shape the rules governing the fast-moving technology. The 28-page AI Action Plan marks US President Donald Trump's administration's first comprehensive strategy on the topic and comes as national security hawks have raised concerns that superior American chips may reach China – including through third countries – and advance its AI and military development. Under the plan, the US Commerce Department will lead an effort to develop new export controls on chipmaking components to close 'loopholes' in current restrictions that now focus on major systems. The department will also lead efforts to increase the monitoring of exported US chip end users and to explore using new chip location verification features to keep them out of 'countries of concern' – a term often used to refer to China, Iran, North Korea and Russia, among others. The plan also directs the Defence and Commerce departments to coordinate with allies on adopting US export controls and to prohibit American adversaries from being involved in their defence supply chains. 'Artificial intelligence is a revolutionary technology with the potential to transform the global economy and alter the balance of power in the world,' White House AI tsar David Sacks said in a written statement. 'To win the AI race, the US must lead in innovation, infrastructure, and global partnerships ... This Action Plan provides a road map for doing that,' he added. The Trump administration decision last week to allow American chipmaker Nvidia to sell its H20 chips to China reinvigorated furious debate about how to extend the US technological lead without compromising national security. That came months after the administration revoked a rule put in place under former President Joe Biden that limited how much US AI computing capability certain countries could access through imported US chips. American AI companies face increased competition from such Chinese rivals as Hangzhou-based DeepSeek, which sent global shock waves earlier this year when it released powerful AI models built far more cheaply than many had thought possible. Trump has previously warned of China's potential to surpass the US in the race for AI dominance. Speaking about the White House plan at an AI summit on Wednesday, Trump said he wants to make the US an 'AI export powerhouse' while maintaining 'necessary protections for our national security'. Commerce Secretary Howard Lutnick and Secretary of State Marco Rubio 'will work to rapidly expand American AI exports of all kinds, from chips to software to data storage', he said, before signing executive orders to fast-track AI development and promote exports. Ryan Fedasiuk, a former technology policy advisor in the State Department's Office of China Coordination, said that taken together, the export control proposals 'thread a difficult needle' in the lead-up to US-China trade talks in Stockholm next week. 'They tighten technology controls without risking an existential blow to the provisional trade deal struck in London, and they focus on enforcement gaps rather than adding sweeping new restrictions,' he said. The real impact of the plan will depend 'entirely' on its execution, Fedasiuk added. 'Allied coordination is where the rubber meets the road. Export controls only work if they're multilateral, and we've seen too many cases where US restrictions just pushed business to Dutch or Japanese competitors.' The new administration blueprint also makes several recommendations focused on countering the spread of Chinese models. To ensure 'that free speech flourishes in the era of AI', the Commerce Department is to evaluate advanced Chinese AI models for alignment with Communist Party talking points and censorship practices. Along with the US State Department, Commerce will lead a campaign to 'vigorously advocate' for international AI governance approaches that 'counter Chinese influence' at diplomatic and standard-setting bodies. Too many existing efforts 'have advocated for burdensome regulations, vague 'codes of conduct' that promote cultural agendas that do not align with American values, or have been influenced by Chinese companies attempting to shape standards for facial recognition and surveillance', the plan said. The two departments will also lead a new initiative to share technology protection measures, including measures in basic research and higher education, with allies and partners. To promote the use of American AI technology rather than that of adversaries, the Commerce Department will gather industry proposals for 'full-stack AI export packages' – bundled systems that include AI hardware, software, and related services intended for export. In recent years, the US has pushed its European and Asian allies to stop exports of advanced chipmaking equipment to China. Wednesday's document carries a thinly veiled threat to allies, warning that failure to align with US export controls could trigger punitive trade measures through secondary tariffs. - SOUTH CHINA MORNING POST
Malaysian Reserve
13 hours ago
- Malaysian Reserve
China urges global consensus on balancing AI development, security
SHANGHAI — China's Premier Li Qiang warned Saturday that artificial intelligence development must be weighed against the security risks, saying global consensus was urgently needed even as the tech race between Beijing and Washington shows no sign of abating. His remarks came just days after US President Donald Trump unveiled an aggressive low-regulation strategy aimed at cementing US dominance in the fast-moving field, promising to 'remove red tape and onerous regulation' that could hinder private sector AI development. Opening the World AI Conference (WAIC) in Shanghai on Saturday, Li emphasised the need for governance and open-source development, announcing the establishment of a Chinese-led body for international AI cooperation. 'The risks and challenges brought by artificial intelligence have drawn widespread attention… How to find a balance between development and security urgently requires further consensus from the entire society,' the premier said. Li said China would 'actively promote' the development of open-source AI, adding Beijing was willing to share advances with other countries, particularly developing ones. 'If we engage in technological monopolies, controls and blockage, artificial intelligence will become the preserve of a few countries and a few enterprises,' he said. 'Only by adhering to openness, sharing and fairness in access to intelligence can more countries and groups benefit from (AI).' The premier highlighted 'insufficient supply of computing power and chips' as a bottleneck. Washington has expanded its efforts in recent years to curb exports of state-of-the-art chips to China, concerned that these can be used to advance Beijing's military systems and erode US tech dominance. For its part, China has made AI a pillar of its plans for technological self-reliance, with the government pledging a raft of measures to boost the sector. In January, Chinese startup DeepSeek unveiled an AI model that performed as well as top US systems despite using less powerful chips. At a time when AI is being integrated across virtually all industries, its uses have raised major ethical questions, from the spread of misinformation to its impact on employment, or the potential loss of technological control. In a speech at WAIC on Saturday, Nobel Prize-winning physicist Geoffrey Hinton compared the situation to keeping 'a very cute tiger cub as a pet'. 'To survive', he said, you need to ensure you can train it not to kill you when it grows up. In a video message played at the WAIC opening ceremony, UN Secretary-General Antonio Guterres said AI governance would be 'a defining test of international cooperation'. The ceremony also saw the French president's AI envoy, Anne Bouverot, underscore the 'an urgent need' for global action. At an AI summit in Paris in February, 58 countries including China, France and India — as well as the European Union and African Union Commission — called for enhanced coordination on AI governance. But the United States warned against 'excessive regulation', and alongside the United Kingdom, refused to sign the summit's appeal for an 'open', 'inclusive' and 'ethical' AI. — AFP
The Sun
a day ago
- The Sun
Tesla tops Chinese rivals in assisted driving tests
BEIJING: Billionaire Elon Musk's Tesla outperformed Chinese rivals including BYD, Xiaomi and Huawei in a test of assisted driving technologies on China's highways, according to results published by TikTok owner Bytedance's auto unit Dcar. State television CCTV and Dcar jointly tested the level 2 advanced driving assistance systems (ADAS) from more than 20 electric vehicle brands in China and rated their performance in a series of scenarios with higher risks of accidents on highways and urban traffics. The test videos posted by Dcar went viral on Chinese social media. Tesla scored the best in the highway test among 36 models, with its Model 3 and Model X passing five out of six scenarios, while BYD's Denza Z9GT and Huawei-backed Aito M9 failed in three scenarios. Xiaomi's SU7 passed in one of six. In a Weibo post on Friday, HIMA, the Huawei-led auto alliance, said it declined to comment on the 'so-called test.' BYD and Xiaomi didn't immediately respond to requests for comment. 'Due to laws against data export, Tesla achieved the top results in China despite having no local training data,' Tesla CEO Elon Musk said on his X account on Friday. Tesla has been caught in what Musk described as a 'quandary', as the U.S. doesn't allow its AI software to be trained in China, while the automaker has been seeking approval from Chinese regulators to transfer data saved locally in Shanghai back to the United States for algorithm training. Domestic brands should face up to the gap with Tesla in autonomous driving, Wang Yao, deputy chief engineer of the China Association of Automobile Manufacturers, told an auto forum in Shanghai earlier this month. Xiaomi CEO Lei Jun, in remarks after a Tesla Model Y delivered itself from an Austin, Texas factory to its owner in the area roughly 30 minutes away, said 'we will continue to learn' from Tesla which has led industry trends. The test came amid growing safety concerns in China about the ADAS after a highway accident involving a Xiaomi SU7 killed three people in March. State media have blamed misleading promotions for resulting drivers' improper uses of the technologies and the authorities have banned the uses of terms such as 'smart driving' and 'autonomous driving' for marketing driving assistance features. The public security ministry said this week that the country will set out legal responsibilities related to the technology that has yet achieved true autonomous driving. Drivers face safety and legal risks if they are distracted in accidents when assisted driving is turned on, the ministry warned. Xiaomi had seen a slump in new EV orders as a consumer backlash began in April following the fatal trash, but the impact seems short-lived, with its new electric SUV receiving exceptionally strong initially orders after it went on sale last month. Tesla's sales of its China-made electric vehicles edged up 0.8% in June from a year earlier, snapping an eight-month losing streak, but they continued to fall on a quarterly basis in the face of lower-cost new models from its Chinese rivals. Tesla's assisted driving suite is available in China for nearly $9,000, while the technology from its local rivals including Xiaomi and BYD is without extra cost, pressuring the U.S. automaker's self-driving future. Tesla's technology approach relies solely on cameras as sensors and artificial intelligence while most Chinese peers including BYD use lidar (light detection and range sensors) additionally to ensure performance. ($1 = 7.1624 Chinese yuan renminbi) - Reuters
