First human brain-to-brain interface allows remote control over the internet, telepathy coming soon

The first human-to-human, brain-to-brain noninvasive interface has been created by researchers at the University of Washington. The system allows one researcher to remotely control the hand of another researcher, across the internet, merely by thinking about moving his hand. The researchers are already looking at a two-way system, to allow for a more “equitable” telepathic link between the two human brains, and the telepathic communication of complex information.

Despite the massive and mostly-not-understood complexity of the human brain, the UW brain-to-brain interface is actually quite simple, relying on tools that are regularly used in the fields of medicine and brain-computer interfaces (BCIs). The first human brain (the sender) is connected to a computer via an EEG-based BCI. The second human brain (the receiver) is connected to another computer via a Magstim transcranial magnetic stimulation (TMS) machine — the same kind of TMS setup that has been somewhat successful in treating depression, and other mental maladies. When the sender plays a game and thinks about firing a cannon at a target, the EEG picks it up, sends the signal across the internet to the second computer, and the TMS stimulates the region of the receiver’s motor cortex that controls hand movement. This causes the receiver’s index finger to twitch, firing the cannon and blowing up the target. This process is almost instantaneous.

TMS is a lot like transcranial direct current stimulation (tDCS), which we have written about extensively. Where tDCS passes an electrical current through your brain, affecting the neurons that the electrons travel through, TMS uses electromagnetic induction to create a similar effect. Both tDCS and TMS can be used to either stimulate regions of the brain, useful for brain-to-brain interfaces or increasing the activity of regions of the brain associated with depression, or to reduce the activity of a region, which might help with the treatment of other conditions, such as Parkinson’s. Like tDCS, TMS is completely noninvasive, and so far it appears to be completely safe.

The University of Washington (UW) researchers, led by Rajesh Rao and Andrea Stocco, have basically connected two quite simple and well-understood systems into a novel and slightly terrifying human-to-human interface. It is very similar to Harvard’s human-to-mouse interface, except they used focused ultrasound (FUS) instead of TMS to trigger the motor cortex. That the UW setup works isn’t all that surprising — the main thing is that that, for the first time, a human is on the receiving end, which raises some interesting ethical and moral issues.

Chantel Prat, another researcher involved with the work, is quick to try and dispel any concerns. “I think some people will be unnerved by this because they will overestimate the technology,” Prat says. “There’s no possible way the technology that we have could be used on a person unknowingly or without their willing participation.” This is an overly simplistic way of looking at it, though. Yes, the current setup requires both users to be fully consenting — but in the future, it’s not hard to imagine wireless implants that allow for full telepathy and perhaps a limited range of remotely triggered actions. (See: Brown University creates first wireless, implanted brain-computer interface.) As always with technology, we don’t need to worry so much about the hardware itself — but rather how it might be subverted, once a significant number of people have brain-to-brain interfaces installed.

Moving forward, Rao and Stocco are now working on transmitting more complex information between two human brains. This could be done fairly simply with encoded pulses — think brain-to-brain Morse code — or they could go the complex route and try to stimulate the brain into creating actual images and thoughts. There’s still a lot of work to be done to decode the human brain, so it will be very interesting to see how future human-to-human brain-to-brain interfaces are implemented.

Curiosity turns on self-driving software, can now navigate Mars on its own

Roughly 200 million miles away on the surface of the Red Planet, NASA’s Curiosity just one-upped Google: It autonomously drove across the surface of Mars without human supervision. This is the first time that Curiosity has turned on its self-driving “autonav” software, and initial reports suggest that Curiosity navigated the treacherous surface of Mars flawlessly. The autonav software will help Curiosity reach its destination, Mount Sharp, much more quickly because the rover won’t have to wait for driving instructions from NASA — it can just plug away at the remaining kilometers autonomously.

Up until yesterday, every movement made by Curiosity has been painstakingly keyed in by NASA, usually after performing simulations here on Earth using Curiosity’s stunt double (theVehicle System Test Bed). These movements are planned by NASA engineers, who pore through photos of the terrain captured by Curiosity to seek out potential obstacles, such as big rocks or sand traps (Mars rover Opportunity famously stumbled into a sand dune in 2005, and took 40 days to extricate itself). The problem is, the cameras on board Curiosity can only see so far ahead; if there’s a dip in the ground, or the rover is going up hill, NASA can only plan a very short drive until it gets updated imagery of the rover’s surroundings.

The view from Curiosity’s front-left hazcam, with Mount Sharp in the distance

This is where the autonomous navigation software, or autonav for short, kicks in. Basically, Curiosity is equipped with two stereo pairs of hazard avoidance cameras (hazcams) which create a 3D map of the rover’s surroundings. This is similar to the EyeSight system implemented by the Subaru Forester, but a lot simpler than the LIDAR system employed by Google’s self-driving cars (Mars doesn’t have any fast-moving obstacles, making autonomous driving a lot easier.) Using this 3D map, the rover can plot an alternate course around any obstacles that aren’t safe to drive over. Yesterday, August 27, Curiosity’s 376th Martian day, the rover autonomously drove itself through a 10-meter depression which NASA could not confirm ahead of time to be safe.

A poster illustrating Opportunity’s autonav software, which is very similar to Curiosity (click to zoom in)

A map of Curiosity’s progress from landing, to Glenelg, to its eventual target: Mount Sharp

Since leaving Glenelg, where Curiosity confirmed that conditions on Mars could’ve once supported life, the rover has driven a grand total of 0.86 miles (1.39 km) towards its primary science target, Mount Sharp. It has around 4.46 miles (7.18 km) left to go, and will stop at a number of scientifically interesting waypoints identified by the Mars Reconnaissance Orbiter’s HiRISE camera. It will take months to reach Mount Sharp, but we should have a lot of pretty photos and interesting science to share during that period.

Honda Fit: 86 mpg from the next hyper-efficient hybrid

Could your next car get 86 mpg? It might if it’s a Honda. The next-generation Honda Fit subcompact will be unveiled this fall and arrive in the US in the first half of 2014. Most of the buzz over the new Fit, called the Honda Jazz in some countries) is the hybrid version, which promises a 35% improvement in fuel economy. The US currently gets the gasoline Honda Fit and EV Fit — not the hybrid Fit — but that could change with the next model.

On a Japanese test cycle, 2014 Fit Hybrid fuel economy will be on the order of 2.7 liters consumed per 100 km or 85.6 US mpg. That’s a mathematical conversion that doesn’t account for the US test cycle. But still, it could be the most efficient hybrid if and when it arrives stateside. Currently the most efficient non-EV cars sold in the US are the Toyota Prius C and Toyota Prius, each with 50 mpg combined EPA rating, 53 mpg and 51 mpg city ratings for the Prius C and Prius, respectively. The 2013 Honda Fit gets 29-31 mpg combined depending on the transmission or 33-35 mpg highway; the Honda Fit EV gets 118 mpg-e (miles per gallon equivalent), best in the category the EPA calls small station wagons.

Atkinson engine, 7-speed double clutch transmission, electric motor

Under the umbrella of Honda’s Earth Dreams Technology program, the new Fit Hybrid will be the first to employ Honda’s Intelligent Dual Clutch Drive system, or i-DCD. The gasoline power comes from a 1.5-liter, four-cylinder Atkinson cycle engine. A single electric motor is packaged with a seven-speed dual clutch transmission, linked through an intelligent power unit (IPU) to a lithium-ion storage battery. An Atkinson engine has an effectively shorter compression stroke (when the piston moves upward) than the downward power stroke, accomplished by not immediately closing each cylinder’s intake valve. It captures more of the power in the fuel-air mixture.

When starting out, the clutches disengage the gas engine and the Fit Hybrid starts off on battery power. The clutches engage the engine and gearbox under sporty (hard) acceleration and at higher speeds. When the Fit Hybrid decelerates, the gas engine is again disengaged.

Honda has claimed a 35% increase is compared to the current Fit Hybrid’s integrated motor assist (IMA) configuration. The IMA electric motor only runs when the gasoline engine runs, functioning much like a turbocharger. This is called a mild hybrid or weak hybrid configuration. A hybrid such as the Prius that can run on battery power alone is a strong hybrid, and who wouldn’t prefer strong over mild, let alone weak?

IMA may be cost-effective tech, but lots of hybrid owners want to see — and show their friends — a car that runs on battery alone for a mile or two. This is what happens when you let engineers have input into how a company runs: They pick solutions that make cost-effective sense. Sometimes the market agrees, other times not. Honda has long argued nobody needs an engine of more than six cylinders and it turns out they’re right, but that hurt the Acura brand competing against V8 offerings from Lexus, Audi, BMW and Mercedes-Benz the last two decades. Honda has also been one or two gears shy of the competition in transmissions, arguing that four or five forward gears was  fine — “look at our mpg figures, not the numbers of gears.” Now, they too are creeping upwards to as many as seven (the industry record is currently nine). The IMA hybrid is giving way to the i-DCD hybrid. At the very least, IMA served its purpose and Honda is moving on to a more efficient technology.

Three hybrid flavors from Honda

Honda will actually have three hybrid configurations going forward with one, two, and three electric motors for small, medium, and large/sporty cars. All will use lithium-ion batteries. The previous IMA hybrids used nickel-metal hydride.

The one-motor (electric motor) Intelligent Dual Clutch Drive design is what’s on the new Fit hybrid and likely other small Hondas. The dual-clutch system is an automated mechanical transmission, meaning there’s a clutch (actually, two) as on manual gearboxes, but it’s automated. The car, rather than the driver’s left foot, activates the clutch. Many European cars use DCT transmissions and they’ve all but killed off the manual gearbox among performance cars.

The two-motor Intelligent Multi Mode Drive promises the highest efficiency, according to Honda researchers. It’s suited for plug-in hybrids, the ones that can go 20-40 miles on batteries before switching over to the combustion engine. It will be on the 2014 Honda Accord plug-in hybrid electric vehicle (PHEV). If there’s enough battery power, this drivetrain offers three modes. EV Mode is battery-only; it regains some of the expended energy braking and going down hills. Engine Drive is for medium and high-speed driving; the hybrid stuff is along for the ride here, with the gasoline engine directly connected by a lock-up clutch to the drive wheels. Hybrid Drive is a combination for city driving and also for extra power accelerating on the highway, where the electric motor adds boost just as a turbocharger does.

The three-motor Super-Hybrid All-Wheel-Drive mode (SH-AWD) is for performance Hondas and Acuras. It uses a V6 gas engine, has the performance of a V8, and the fuel economy of a four-cylinder, Honda says. A 3.5-liter V6 engine is up front, along with the seven-speed DCT and an electric motor to drive the front wheels. Two more electric motors in back provide power for the left and right rear wheels and torque distribution. There is no driveshaft sending mechanical power to the rear wheels, which saves space and weight. Acura’s existing SH-AWD gas-engine cars have the ability to overpower the outside wheel going around corners or on slippery patches. Depending on what kind of driver you are, SH-AWD provides an added measure of safety or performance.

Cellphone use doesn’t increase the number of car accidents, says new Carnegie Mellon study

Cops may still write you a ticket for yakking on a handheld phone while driving. But the link between cellphone use and accidents looks more tenuous if you agree with the conclusions of a recent study from Carnegie Mellon University and the London School of Economics and Political Science. Using statistics and data comparisons, the researchers found that the increased use of cellphones has led to no measurable increase in accidents. Expect this study to be hotly contested. It flies in the face of conventional wisdom and the prevailing winds in Washington that would like to see more restrictions on phone use and texting, up to and including interlocks that keep anyone in the car from using their phone.

The study was published in the American Economic Journal: Economic Policy. Saurabh Bhargava, assistant professor of social and decision sciences in CMU’s Dietrich College of Humanities and Social Sciences and Vikram S. Pathania of the London School of Economics and Political Science looked at cellular data from 2002 to 2005, apparently after the NSA was done parsing it. They identified drivers as those whose calls were regularly handed off from cell tower to cell tower. At the time, most carriers offered free calling after 9 pm. Bhargava and Pathania found motorists increased their calling by 7% at 9pm. They pulled up data on eight million car crashes in those four years as well as all fatal crashes. Their finding: no statistical link. Crashes didn’t go up when calling went up.

As a bonus, they compared states that enacted handheld cellphone bans and found no difference in the crash rate before and after. There’s a damning graphic in their study that compares cellphone ownership over 20 years vs. car crashes. Ownership is steadily up, crashes and fatal crashes are steadily down.

Are they on to something, or just blowing smoke?

It’s easy to challenge the study, or at least to nibble around the edges. The mobile user data would include passengers as well as drivers. Cars are getting safer all the time and drunks are being policed off the roads more than in the past. Passing a hands-free law isn’t the same as getting people to obey it. Texting may be more distracting than making a call, as the authors themselves note. And so forth. But still, the research should be kept in mind when the the Department of Transportation and the states ponder their next steps.

“Using a cellphone while driving may be distracting, but it does not lead to higher crash risk in the setting we examined,” says Bhargava. “While our findings may strike many as counterintuitive, our results are precise enough to statistically call into question the effects typically found in the academic literature. Our study differs from most prior work in that it leverages a naturally occurring experiment in a real-world context.”

Statistical modeling of big data can be a powerful tool to prove something that would be difficult otherwise. One example: Are blacks and Hispanics on death row proportionally more than whites because they’re inherently, ah, criminalistic? Work with a big data sets showing arrest, trial, conviction, and sentencing for whites from similar socioeconomic backgrounds and similar crimes, and many researchers say the one variable that best explains the discrepancy is the defendant’s color. Conservatives and traditionalists may not like statistics and deep research to prove a point. But Nate Silver didn’t pick all 50 states right in the 2012 presidential election by flipping a quarter.

Cellphone-in-car opponents mostly rely on a 1997 article in the New England Journal of Medicine. It said cellphone use by drivers increased the risk of a crash fourfold, making it a safety hazard on par with driving drunk.