Robot, dress thyself
It may have no need of clothing, but a robot that can slip on a T-shirt could go a long way toward building automatons that help humans with everyday tasks. Typically, robots are great at grasping solid, regular-shaped objects 鈥� something made of cloth, not so much. No more. Takamitsu Matsubara of the Nara Institute of Science and Technology in Japan built a robot that uses machine learning to figure out how to orient the floppy garment and put its arms through the correct holes (). The ability could be used to build a robot that can apply bandages or help a person get dressed, says Matsubara.
Your phone knows where it is
Advertisement
Can鈥檛 find your phone anywhere? Life would be much easier if your phone could help. Now software can give phones the ability to know where they are, making them easier to locate when misplaced. Developed by Jason Wiese and researchers at Microsoft, the Phoneprioception system uses a phone鈥檚 accelerometer and light sensor to determine whether it鈥檚 in your pocket, say, or lying in the bottom of a bag. Users could then log onto another device and use a 鈥渇ind my phone鈥� service to access this information.
Tweets reveal postnatal depression
Though usually a happy event, having a baby can trigger postnatal depression for some new mums. Now tweets can be used to predict which women are likely to develop the condition. Munmun De Choudhury of Microsoft Research and colleagues analysed the tweets of 376 expectant mums. By tracking several factors including emotional content and the linguistic style of their tweets, the team predicted which mothers would suffer an extreme negative shift in behaviour after giving birth 鈥� potentially a sign of depression 鈥� with 71 per cent accuracy. The system was presented at the Human Factors in Computing Systems conference in Paris, France, last week.
