could help people with very limited vision. Researchers at Oxford University are developing glasses that overlay images of nearby objects on the see-through lenses. The idea is to

the wearer already has. The glasses include a camera that feeds data to a small processing unit and software that sends images back to the glasses. The software interprets nearby surroundings and overlays shapes of items such as kerbs, tables and chairs, or groups of people on what the wearer can already see. Further development may include facial, object or text recognition, with audio supplied to an earpiece. This suggests interesting possibilities for specialised driving glasses.

FAMILIAR FACES: Rare genetic diseases may affect around 6% of the population, but they can be tricky to diagnose. There are genetic tests for some, such as Down's syndrome, but many haven't yet been traced back to gene variants. Facial features can often provide a clue though, provided you know what to look for, which most doctors don't. That's where facial recognition software can come in. Researchers in the UK have developed software that scans full-frontal facial photos captured with a cellphone or other camera. The software analyses the picture and produces a description of the face which it then compares to a database of faces of people with known disorders. Ultimately it produces a ranking of possible disorders to investigate. The software recognises 90 disorders and tests have shown it to be around 93% accurate. The software could be very useful as a first step in a diagnostic process, especially in countries where genetic screening isn't available. That's smart: a doctor shouldn't have to know things software could do better, such as matching faces to a database.


A NEW SORT OF THINGS: Do you really care about the most efficient way to sort a large number of things? You might if that sorting would detect a few tumour cells in a large quantity of biological fluid. Currently analysing samples takes hours or days, but a new inertio-elastic flow focusing sorting technique developed by MIT could mean test results at the bedside in real time. For example, a few litres of fluid drained from a patient's lungs could contain millions of cells, including some from tumours. Locating those tumour cells could allow a physician to diagnose cancer. Researchers sent biological samples through the channels of a microfluidic device and found ways to adjust the flow so all the larger particles were concentrated at the centre. A high-speed, pulsed-laser imaging system took snapshots of the shapes, sizes, and orientations of the particles as they flew through the device. Ultimately a device like this could be useful in medicine, manufacturing or perhaps water purification. Finding something is always quicker of you know roughly where to look in the first place.

BUILDER BOTS: It's already been demonstrated that a 3D printer can be used to build a house, but the problem is the printer needs to be bigger than the structure it's creating. A team at the Institute for Advanced Architecture in Catalonia have developed Minibuilders: autonomous bots that divide up tasks to make the construction process cheaper and greener. The bots can work together to build structures of any size, and that don't require support structures either. At the foundation level robots move in a track to squirt out material that hardens into the shell of the building. Then robots clamp on to the existing structure to build out more layers, frames for windows and doors, and ceilings. Finally vacuum robots add a reinforcing layer. The developers have demonstrated the process as a proof of concept, but now it will need to be scaled up and applied. Security and programming flaws could be the downfall of robot building.

ADDING TO SUBTRACT: Various industries, including those in transport, need materials that are light, stiff and strong. US scientists have created one such material through additive micro-manufacturing processes. Their micro-architected metamaterials maintain a nearly constant stiffness per unit mass density, even at ultralow density. That could make them useful in parts and components for aircraft, automobiles and space vehicles. The lightweight materials can withstand a load of at least 160,000 times their own weight because of their geometric layout at the microscale rather than their chemical composition. The team was able to build microlattices out of polymers, metals and ceramics. While others have created ultra-lightweight lattice materials before now, these are 100 times stiffer. Shaving weight off planes and other vehicles can translate to considerable fuel savings and reduced emissions.

Miraz Jordan,