Brazil has a big electricity theft problem. Who is responsible for the theft in Brazil isn't always clear – sending meter readers to check whether meters and overhead cabling have been tampered with is dangerous work, says Adrian Grilli of the Joint Radio Company in London, which does telecommunications for global energy companies. The most accurate versions of the system were able to identify problem cases just over 65 per cent of the time, which the team believes outperforms similar tools. Similar false positives may also be leading to erroneous billing by energy companies.
A 2016 study by the US Bureau of Justice Statistics found that police records cover half as many police-related deaths as media reports. His team's algorithm analyses sentences in news reports to try to extract the names of people who have been killed. An algorithm extracted sentences that it thought referred to police shootings, and then compared these sentences to the names in Burghart's database to find out which ones referred to people who had actually been shot by police. Despite the narrow focus, Burghart says machine learning will eventually mean that he doesn't have to spend hours every week digging through past news reports to pull together records of police-related deaths.
FARMERS can now zap their crops with a handheld scanner to instantly determine nutritional content, which could prove crucial in mitigating the effects of climate change on food quality. "Real-time results mean farmers can add fertilisers or tweak moisture levels as crops grow" Farmers can use the app to assess the impact of changing conditions, such as extreme weather and soil quality, on the quality of their crops from year to year. It could allow farmers to mitigate the negative effects of climate change early by adding fertilisers or tweaking moisture levels as crops grow. Other companies are developing similar gadgets for consumers, and sensors that can be fitted onto a smartphone.
Aggression and sexual behaviour are controlled by the same brain cells in male mice – but not in females. The brain regions that contain these cells look similar in mice and humans, say the researchers behind the study, but they don't yet know if their finding has relevance to human behaviour. They discovered a set of cells within this region in male mice that controlled both aggressive and sexual behaviours. Plus other regions that are known to look different in male and female mouse brains have "considerable overlap" in the brains of women and men, she says.
Could chatbots lend a non-judgemental ear to people making decisions about the end of their life? The earlier people start considering how they want to die and what they want to happen afterwards, the easier it is for those around them to act on those decisions – for example, ensuring they don't die in hospice if they would prefer to be at home. Writer and film-maker Avril Furness agrees that technology can be a useful way to help people start having difficult conversations about death. Furness, who has explored the subject of assisted suicide, says Bickmore's chatbot system is another good way to get people thinking about the end of their life, helping them work through their feelings without worrying what someone else thinks.
Developed by the London-based HR company Saberr, it asks about workplace dynamics and provides the team with reports. A unit within the UK's National Health Service is trialling it, as are 10 companies, including Unilever and Logitech. "Team members start by saying hello to CoachBot, and are then asked about who they are and what they do," says Tom Marsden, Saberr's CEO. Questions Coachbot asks include "Is your team productive?"
Artificial intelligence can identify changes in the brains of people likely to get Alzheimer's disease almost a decade before doctors can diagnose the disease from symptoms alone. This was intriguing, says La Rocca, since this is similar to the size of the anatomical structures connected with the disease, such as the amygdala and hippocampus. "Nowadays, cerebrospinal fluid analyses and brain imaging using radioactive tracers can tell us to what extent the brain is covered with plaques and tangles, and are able to predict relatively accurately who is at high risk of developing Alzheimer's 10 years later," says La Rocca. In contrast, the new technique can distinguish with similar accuracy between brains that are normal and the brains of people with MCI who will go on to develop Alzheimer's disease in about a decade – but using a simpler, cheaper and non-invasive technique.
However, Apple are only talking about one type of error that could happen with its Face ID algorithm – someone else gaining access. "The two errors are related; for a given biometric system, if you reduce one type of error, the other type of error goes up," says Anil Jain at Michigan State University. Last year, police asked Anil Jain at Michigan State University to help unlock a murder victim's Samsung Galaxy S6 phone. True face recognition involves matching a face to a vast database of faces, like police trawling CCTV images for potential suspects.
People often don't realise that they're not adopting the right posture when lifting heavy items, says Eya Barkallah at the University of Quebec at Chicoutimi in Canada. This combination of sensors is perfect for spotting a lot of posture problems, says James Brusey at Coventry University, UK, who wasn't involved in the study. The team at Quebec had a volunteer put on the hat and shoes and lift some boxes in three different ways – half of the time the volunteer used best practice, but the other half, they deliberately lifted while making the most common lifting mistakes. But Subramanian Ramamoorthy at the University of Edinburgh, UK, thinks there are better ways to work out whether someone is moving in the right way.
Amarjot Singh at the University of Cambridge and his colleagues trained a machine learning algorithm to locate 14 key facial points. The researchers then hand-labelled 2000 photos of people wearing hats, glasses, scarves and fake beards to indicate the location of those same key points, even if they couldn't be seen. The system accurately identified people a wearing scarf 77 per cent of the time – a cap and scarf 69 per cent of the time and a cap, scarf and glasses 55 per cent of the time. Last year, a team of researchers from Carnegie Mellon University found they could trick face recognition software by wearing specially designed glasses.