In Brief:
A London-based startup has combined some of today’s most disruptive technologies in a bid to change the way we’ll build the future. By retrofitting industrial robots with 3D printing guns and artificial intelligence algorithms, Ai Build has constructed machines that can see, create, and even learn from their mistakes.
When CEO and founder Daghan Cam was studying architecture, he noticed a disconnect between small-scale manufacturing and large-scale construction. “On one side we have a fully automated production pipeline,” Cam explained at a recent conference in London. “On the other side we’re completely dependent on human labor.” With the emergence of more efficient printing technologies, he thought there must be a better way.
Continue reading “Give a 3D printer artificial intelligence, and this is what you’ll get” »
If you don’t like the thought of bugs crawling all over you, then you might not like one possible direction in which the field of wearable electronics is heading. Researchers from MIT and Stanford University recently showcased their new Rovables robots, which are tiny devices that roam up and down a person’s clothing – and yes, that’s as the clothing is being worn.
The centimeter-sized robots hang on by pinching the fabric between their wheels, with the physically-unconnected wheel on the underside of the material held against the others simply by magnetic attraction.
While self-driving vehicles will be revolutionary, having self-driving vehicles communicate with one another and with road infrastructure will take that revolution to the Nth degree. In a UK-first, a cross-company collaborative demonstration of these technologies working together has taken place.
Researchers at Microsoft have published details of new speech recognition technology that they say transcribes conversational speech as well as a human does. “We’ve reached human parity,” says Microsoft’s chief speech scientist Xuedong Huang in a statement. “This is an historic achievement.”
The system’s word error rate is reported to be 5.9 percent, which Microsoft says is “about equal” to professional transcriptionists asked to work on speech taken from the same Switchboard corpus of conversations. It uses neural language models that group similar words together, allowing for efficient generalization. Microsoft plans to use the technology in Cortana, its personal voice assistant for Windows and the Xbox One, as well as speech-to-text transcription software.
Although the results are impressive, it’s far from an endgame for speech recognition. Microsoft still needs to tune the technology to work as well with conversations in a wider range of more challenging real-life situations and with a broader selection of voices. And for use cases such as Cortana, much of the difficulty comes from teaching the artificial intelligence to understand the meaning of words and act on them, not just accurately hear them.
Continue reading “Microsoft reaches ‘human parity’ with new speech recognition system” »
I know, it doesn’t seem like there’s any possible way that a transmission system could be interesting enough that we’d dedicate an entire article (and video!) to it. But here we are: As soon as SRI explained how their new Abacus transmission worked, we were absolutely sure that it was cool enough to share. In a nutshell, here’s why: It’s the first new rotary transmission design since Harmonic Drive introduced its revolutionary gear system in the 1960s*, and it might give harmonic gears a literal run for their money.
The physics of most electric motors generally dictates that the motors are happiest when they’re spinning very fast. Unless you want to use them to simply spin a thing very fast, you’ll need to add a rotary transmission that can convert low torque, high speed rotation into higher torque, lower speed rotation. If you’ve got the budget, the way to do this is with a high-performance harmonic gear like the ones offered by Harmonic Drive. Roboticists like harmonic gears because they are compact, have high gear ratios, and, perhaps most important, don’t have backlash, which is essentially the amount of wiggle room that you get with conventional gear-based transmissions. In robotic applications, wiggling means that you don’t know exactly where everything is all the time, making precision tasks something between irritating and impossible.
Harmonic gears are great, but they’re also superduper expensive, because they require all kinds of precision machining. Alexander Kernbaum, a senior research engineer at SRI International, has come up with an entirely new rotary transmission called the Abacus drive, and it’s a beautiful piece of clever engineering that offers all kinds of substantial advantages:
Continue reading “SRI Demonstrates Abacus, the First New Rotary Transmission Design in 50 Years” »
Concerns that artificial intelligence will pose a danger if it develops consciousness are misplaced, says Jim Davies.
The REAL danger of AI… wink
Robots (Emily Blunt, Mikey Day) meant to deliver food repeatedly malfunction during a presentation.
