Thanks to researchers from Northwestern Polytechnical University, robot guide dogs that use an artificial intelligence language model may soon be accessible to 17 million people with visual problems in China. The AI guide dog, developed by the university, can assist with tasks like crossing streets, using elevators, and navigating (导航) indoor spaces without being connected to the Internet.

Robot guide dog is an intelligent machine with a physical form. It can interact with the environment in real-time using sensors to execute instructions or make decisions. “Though in recent years there have been some electronic navigation aids, such as electronic navigation helmets, and electronic guide dogs, they cannot engage in conversations with people or fully understand human instructions,” said Sun Zhe, an associate professor at the university.

“They can only carry out tasks according to pre-set programs. They are just cold machines. That’s not enough. The research makes it possible to provide better guidance services for people with visual problems, especially for emotional companionship, ” Sun said.

“People with visual problems face numerous challenges in daily life, including navigation difficulties and safety threats,” Sun said. “Smart guide dogs using the language model can offer them more convenient and safe navigation, effectively improving their quality of life.”

Training a traditional guide dog can cost as much as 200,000 yuan ($ 28,000), and there are fewer than 200 guide dogs in China. Therefore, the research team began developing the smart guide dog based on the AI model, which is much more affordable for people with visual problems. Its findings have been published in the Communications of the Chinese Computer Society.

“The current results prove the possibility of this approach, but there is still some wayto go before it can be put into practical application,” Sun said. “We will put more effort into it and make it available for those in need as soon as possible.”

Scientists have found that floating solar panels could provide a huge amount of electricity if they were placed on lakes and other bodies of water around the world. Floating solar panels could also help save water and protect land.

Solar panels — also known as “photovoltaic” panels — are used to turn sunlight into electricity. Most solar panels are placed on land in large collections called solar farms. But recently people have begun to explore putting floating solar panels on water. Because these panels float, some people call them “floatovoltaies (浮动光伏)”.

The researchers behind the new study looked at 114,555 reservoirs (水库) worldwide. They used computer programs to figure out how much electricity could be produced yearly by covering 30% of these reservoirs with floating solar panels. The answer was surprisingly large — more than twice the amount of energy the United States generates in a year. And 10 times as much energy as all the solar power currently being generated in the world. The researchers described the results as “remarkable”.

The scientists found that floatovoltaies would be especially useful when reservoirs were near smaller cities (50,000 people or less). The researchers say there are about 6,256 cities around the world where floating solar panels could provide all the electricity the cities need.

Floatovoltaies can also help save water by limiting evaporation (蒸发) from reservoirs. The scientists say that solar panels covering just 30% of the reservoirs’ surfaces could save as much water as 300 million people would use in a year. There are several other reasons that make floatovoltaies a good idea. Photovoltaic panels work better when they’re not extremely hot. The water helps cool the panels so that they create more energy. Putting solar panels on water also means that there’s no need to clear land for a solar farm. That’s more and more important as countries work to fight climate change and protect natural spaces for wild animals.

Vocal cords (声带) can easily get damaged by stress, infections, or overuse. It is not just overenthusiastic performers who are at risk of injuring their voice-boxes — according to a study, 30% of the population will experience a voice disorder at some point in their life.

In a study published in the journal Nature Communications, a team of UCLA engineers have invented a patch-like (片状) device that can be worn on the skin outside the throat to help people with voice disorders speak.

Even though the team’s device is an early prototype, it has the potential to offer a substantial improvement on current alternatives. When a person loses his voice today, the easiest fix is to resort to typing, texting, or writing notes to communicate. Typing can be slow and inconvenient, and writing notes is only possible in good lighting. More advanced solutions require special training to use, and surgical interventions are often invasive. A patch would theoretically be able to clear all these hurdles.

The tiny patch consists of a self-powered sensing component, which detects and changes signals generated by the throat muscle movements into high-accuracy, analyzable electrical signals, which are then translated into speech signals using a machine-learning algorithm, and an actuation component, which turns those signals into voice expression. The patch was tested on eight healthy adults who pronounced five sentences (including: “Hope your experiments are going well!”, “Merry Christmas!” and “I love you!”) aloud or in silence, and its accuracy was more than 90%.

There is a way to go yet. For now the device can only recognise the five phrases it was trained on. Plus, individual differences in vocal cords means the algorithm has to be personalised to each user. To make it practical at scale, the researchers will need to collect a lot more data.