Could this be the first smart car for quadriplegics?

In a time where the driverless automated car is becoming a modern reality, we are provided with great potential to make things previously improbable if not impossible suddenly possible.

My interest was piqued when I came across design plans for car that could be controlled by a driver with quadriplegia. At first the idea seemed mere fantasy but as I spoke to transport designer Rajshekhar Dass and learnt move about the control of technical devices through brain waves, facial gestures and infinitesimal movements the idea seemed more of conceivable.

Rajshekar Dass is a car designer from Mumbai, India currently based in Turkey. He has an impressive design history which includes fronting a winner team of the 2016 Michelin Challenge Design for the Google Community Vehicle and  Winner of the  Vehicle, Mobility and Transport Design 2014-15 A’Design Award for a Micro Taxi among awards. He’s interned for Volkswagen in Germany and worked in a range of car dealerships, so has seen car technology from a range of angles.

He detailed his rationale for a car designed specifically for people with quadraplegia, a cohort of people who until now have only featured in the automated cars of the future as mere passengers. Such a design is the first of its kind:

“Audric Design basically was designed with a particular person in mind, Sam Schmidt, former Indy Racing League driver. He was made paraplegic due to a racing car accident in 2000.  He wanted to get back on the track but came back as an owner rather than driver. Everyone loves driving so my interest was how can we use today’s technology to solve these problems for an audience that are generally overlooked? How can today’s tech be used to enable the same driving experience that he enjoyed previously?”

It’s always interesting to learn how a designer approaches the design experience, Dass revealed:

“the first point of research is the capabilities of the human body when functionality has been impaired. In paraplegia the brain signals do not reach the human organs  bellow the neck, so the brain signals are basically lost. I thought, what if you could use today’s technology to enable the signals to be transferred to the computers onboard of a car instead, so you can give a rebirth to the whole driving experience?”

Dass explained that emerging technology like movement through brain wave signals,  gesture and facial recognition and sensor technology s along with  augmented reality made his design more than a well intentioned concept.

There’s precedence here. For example, in 2010  Emotiv released the Emotiv Epoch+, a commercial wearable device designed to enable users to play computer games on a screen through functioning as a brain-computer interface device (Admittedly the design was not without it’s challenges as this review attests).

Then in 2014, Ian Burkhart became the first paralyzed person to use neural bypass technology to pick up and hold a spoon using his own brainpower, with his abilities increasing overtime.

How could it work?

Dass stresses that the car would be completely autonomous in the first instance. But over time, the driver’s gestures and motions would be recorded by the car; for example, as the car is taking a right turn, the driver might be tilting his head. In this way, the car is learning from the driver instead of the driver learning from the car.

Dass explains further:

“There would be a series of levels which would be detected by the AI in the car, that slowly give the control to the driver. e.g. starting with audio and air-conditioning controls first. The next level could be controlling a little bit of the motion and slowly you’d graduate levels as you would in a game with the help of gesture recognition, eye movements and brain mapping. Once the car is confident as to the skill of the driver the complete controls would be available to the driver. but at the same time the car would still have the control over all the systems, as a built-in safety feature.

I’m aware that many people with paralysis experience involuntary movement such as spasms and jerking. I wondered if a car could be smart enough to distinguish these movements from voluntary actions.

Dass agreed:

It’s a good point. The car is completely autonomous and the AI is constantly monitoring the driver’s motions to learn his/her actions, and since the AI is specially developed for paralyzed drivers it can recognize such involuntary movements. Since the AI is also scanning the brain and can understand that the motion has no connection with the brain signals it can be tagged as involuntary action and not require a reaction”.

As well as its driving capabilities, the car would also be designed specifically with the needs of the driver in mind with a rear entry door suited to a wheelchair which would be specially designed to become the driving seat in the vehicle. Dass explained that the project has only been recently published online and he is keen to explore his ideas further with people with disabilities and associated organizations to enable further development.

In an era where ideas as seemingly bizarre as Google’s patent for“sticky” technology to protect pedestrians if they get struck by Google’s self-driving cars, a mind powered car doesn’t seem all that strange at all.

Dass is working on a range of diverse projects currently and judging by the ingenuity inherent in his design portfolio, this is simply an example of things to come.

Could this be the first smart car for quadriplegics?

In a time where the driverless automated car is becoming a modern reality, we are provided with great potential to make things previously improbable if not impossible suddenly possible.

My interest was piqued when I came across design plans for car that could be controlled by a driver with quadriplegia. At first the idea seemed mere fantasy but as I spoke to transport designer Rajshekhar Dass and learnt move about the control of technical devices through brain waves, facial gestures and infinitesimal movements the idea seemed more of conceivable.

Rajshekar Dass is a car designer from Mumbai, India currently based in Turkey. He has an impressive design history which includes fronting a winner team of the 2016 Michelin Challenge Design for the Google Community Vehicle and  Winner of the  Vehicle, Mobility and Transport Design 2014-15 A’Design Award for a Micro Taxi among awards. He’s interned for Volkswagen in Germany and worked in a range of car dealerships, so has seen car technology from a range of angles.

He detailed his rationale for a car designed specifically for people with quadraplegia, a cohort of people who until now have only featured in the automated cars of the future as mere passengers. Such a design is the first of its kind:

“Audric Design basically was designed with a particular person in mind, Sam Schmidt, former Indy Racing League driver. He was made paraplegic due to a racing car accident in 2000.  He wanted to get back on the track but came back as an owner rather than driver. Everyone loves driving so my interest was how can we use today’s technology to solve these problems for an audience that are generally overlooked? How can today’s tech be used to enable the same driving experience that he enjoyed previously?”

It’s always interesting to learn how a designer approaches the design experience, Dass revealed:

“the first point of research is the capabilities of the human body when functionality has been impaired. In paraplegia the brain signals do not reach the human organs  bellow the neck, so the brain signals are basically lost. I thought, what if you could use today’s technology to enable the signals to be transferred to the computers onboard of a car instead, so you can give a rebirth to the whole driving experience?”

Dass explained that emerging technology like movement through brain wave signals,  gesture and facial recognition and sensor technology s along with  augmented reality made his design more than a well intentioned concept.

There’s precedence here. For example, in 2010  Emotiv released the Emotiv Epoch+, a commercial wearable device designed to enable users to play computer games on a screen through functioning as a brain-computer interface device (Admittedly the design was not without it’s challenges as this review attests).

Then in 2014, Ian Burkhart became the first paralyzed person to use neural bypass technology to pick up and hold a spoon using his own brainpower, with his abilities increasing overtime.

How could it work?

Dass stresses that the car would be completely autonomous in the first instance. But over time, the driver’s gestures and motions would be recorded by the car; for example, as the car is taking a right turn, the driver might be tilting his head. In this way, the car is learning from the driver instead of the driver learning from the car.

Dass explains further:

“There would be a series of levels which would be detected by the AI in the car, that slowly give the control to the driver. e.g. starting with audio and air-conditioning controls first. The next level could be controlling a little bit of the motion and slowly you’d graduate levels as you would in a game with the help of gesture recognition, eye movements and brain mapping. Once the car is confident as to the skill of the driver the complete controls would be available to the driver. but at the same time the car would still have the control over all the systems, as a built-in safety feature.

I’m aware that many people with paralysis experience involuntary movement such as spasms and jerking. I wondered if a car could be smart enough to distinguish these movements from voluntary actions.

Dass agreed:

It’s a good point. The car is completely autonomous and the AI is constantly monitoring the driver’s motions to learn his/her actions, and since the AI is specially developed for paralyzed drivers it can recognize such involuntary movements. Since the AI is also scanning the brain and can understand that the motion has no connection with the brain signals it can be tagged as involuntary action and not require a reaction”.

As well as its driving capabilities, the car would also be designed specifically with the needs of the driver in mind with a rear entry door suited to a wheelchair which would be specially designed to become the driving seat in the vehicle. Dass explained that the project has only been recently published online and he is keen to explore his ideas further with people with disabilities and associated organizations to enable further development.

In an era where ideas as seemingly bizarre as Google’s patent for“sticky” technology to protect pedestrians if they get struck by Google’s self-driving cars, a mind powered car doesn’t seem all that strange at all.

Dass is working on a range of diverse projects currently and judging by the ingenuity inherent in his design portfolio, this is simply an example of things to come.

How Cellphone Camera Images Can Fool Machine Vision

Finding some grainy imperfections in your smartphone photo is an unavoidable reality of digital photography (especially in low light conditions), but it’s not going to stop you from recognizing who or what you’ve photographed. However, that might not be true of machines that use Google’s computer vision software to “see.”

According to a new report, Google researchers found that the accuracy of the company’s image recognition algorithms often failed when they were challenged with grainy, less-than-perfect pictures.

We’re already using machine vision for all sorts of purposes, including facial recognition, image identification, and self-driving vehicles. This study looked at Google’s software specifically, which is one of the best machine vision systems out there. It suggests that there could be real limitations to a growing number of such systems, which is important to deal with as we decide how much to trust the devices that technology makers claim can “see” for us.

Adversarial images aren’t a new problem in the field of machine learning. These pictures, which have a specifically engineered type of grainy noise, have been used to throw a wrench into image classification software: Changes to an image that would be just about imperceptible to a human eye, like blurry pixelation, can totally mess up a computer’s ability to correctly identify what it is.

“It was found out, a few years ago, that it is possible to modify the input image and it will confuse the image recognition system,” said Alexey Kurakin, one of the authors of the report and researcher at Google Brain, who spoke to me over the phone from California.

“Let’s say an image of an elephant,” he continued. “You modify the image slightly with this noise that is hard for the human eye to see very well. [If] you give it to your image recognition system, now the image recognition system thinks it’s no longer an elephant, but an airplane or a car,” even though a human eye likely wouldn’t be confused by the same trick.

Before now, this vulnerability had only been tested by uploading an image to the classification system directly. Kurakin and his team tried something different. They took cellphone pictures of printed images that were increasingly modified with that special kind of noise—random data that, again, would not stop us from seeing an elephant, but would throw the computer vision system for a loop.

He and his team found that the software still misclassified items, in the most noisy cases as often as 97 percent of the time.

According to the paper, the cellphone images, which were input into the Inception 3 neural network (which is Google’s really, really smart image identification algorithm), were captured “without careful control of lighting, camera angle, distance to the page.”

In other words, these images looked a whole lot like what would be produced not in the lab, for research purposes, but out in the real world.

“Prior to my paper, they directly fed the image to neural network,” said Kurakin. “This is important because, if you have a file with the image, you have fine-grain control over each pixel.” In other words, previous research had generated the noise in the individual pixels and then let the machine analyze the picture.

Kurakin and his team have now shown that a snapshot from a regular camera with no modifications has the same worrisome effect.

To get over this hurdle, scientists will need to tinker with both the image recognition software itself, and the data that’s used to train it. Maybe machines can be become familiar with flawed images that way.

Until then, it’s a reminder that while we begin to realize the incredible promise of new machine learning technology, it still has fundamental weaknesses that could be exploited.

Intel readies chip to rival NVIDIA for machine learning

After abandoning its own GPU for supercomputers, machine learning, and video games in 2009, Intel has returned to the market with a new 72-core Xeon Phi, to compete with NVIDIA’s growing portfolio of GPUs.

The Xeon Phi ‘Knights Landing’ chip, announced at the International Supercomputing Conference in Frankfurt, Germany last week, is Intel’s most powerful and expensive chip to date and is aimed at machine learning and supercomputers, two areas where Nvidia’s GPUshave flourished.

Inside the chip there is 72-cores running at 1.5GHz, alongside 16GB of integrated stacked memory. The chip supports up to 384GB of DDR4 memory, making it immensely scalable for machine learning programs.

At the conference, Intel mentioned some of the issues with GPUs for complex machine learning programs. It believes that the Xeon Phi, which is a byproduct of the failed GPU in 2009 called Larrabee, is an answer to some of those problems.

Intel has already deployed the new Xeon Phi chip to several supercomputers including the Stampede 2, an 18-petaflop supercomputer that ranks in the top ten worldwide.

While supercomputers are definitely the focus, Intel does see the Xeon Phi chips being utilized by machine learning and artificial intelligence developers. Servers are another area where Intel is looking into, though the $6,294 price tag may put some companies off.

Intel could have an edge over Google

Google has built its own chip, a Tensor Processing Unit (TPU), specifically for machine learning and deep neural networks. While it may be hard for a general chip that doesn’t offer specific functionality to compete with a chip designed for machine learning, some developers may choose Intel, a neutral company, over Google who may be a competitor in the machine learning and AI markets.

Intel has plans to launch an even faster version of the Xeon Phi, clearly showing a commitment to the market that they didn’t have in 2009. Hopefully, if the first Xeon Phi doesn’t succeed, it won’t back away from those plans and give the market entirely to NVIDIA.

Google opens Machine Learning Research Center in Europe to further explore AI

If there was any doubt that artificial intelligence is the future of technology, look no further than Google. It's making a major commitment to Machine Learning (perhaps the only form of AI that matters), by dedicating a Zurich-based research group to it.

The company announced the new AI research push on Thursday in a blog post.

Opening as part of Google's existing Google Research center in Europe, the Machine Learning Research Group will focus on, naturally, Machine Learning, in which computers use vast amounts of data to teach themselves and build rules about the data; Natural Language Processing for speech-systems and conversational queries; and Machine Perception, which is used to understand the contents of images, sounds, music and video.

This announcement comes four years after the company began aggressively pursuing machine learning technologies, three months after Google's AI beat one of the world's best Go players and just weeks after Google codified its approach to AI during its annual Google I/O developers conference.

During I/O, Google unveiled Google Assistant, a voice-based digital assistant designed the take on Microsoft's Cortana, Apple's Siri and Amazon's Alexa. At the time, Google CEO Sundar Pichai said, "Our ability to do conversational understanding is far ahead of what other assistants can do."

Now the creation of a facility dedicated to machine learning is also a signal to competitors like Facebook, which is also making a heavy investment in AI.

And why did Google choose Zurich, as opposed to its Mountain View, California, headquarters? In the blog post, the company points out that the area is home to many of the world's leading technical universities, noting: "We look forward to collaborating with all the excellent computer science research that is coming from the region."

Lutron lights up Amazon Echo’s life

Lutron Electronics has added its Caséta Wireless system to the growing list of devices that support Amazon Alexa range, which includes the Echo, Dot, Tap, and Fire TV.

The Caséta Wireless system features various connected LED lights that can change color or brightness on demand. The integration of Alexa allows users to change the lights through voice commands, an ideal solution when you cannot find your smartphone.

See Also: Could smart cities improve indoor air quality with home sensors?

Amazon has been steadily improving the range of voice commands for the Echo, to the point it can now hear you from a few meters away. So if you want to turn the lights on when you enter your house late at night, there’s a good chance your Echo will hear you from the porch.

Another day, another smart home system looking to break into the market. This time, a team of ex-ASUS engineers have launched a Kickstarter for SmartAll, an AI butler that connects to smart devices.

The hub, which is on sale at Kickstarter, is a small camera with an LED strip around it. It is able to recognize up to five family members and customize a room to their liking — it is also able to spot intruders and create a “panic mode” where the lights and alarms turn on. Users are able to livestream the camera’s POV from a mobile device and save footage. On top of all the camera capabilities, the hub also understands voice commands, like “turn on the television” or “turn up the heating”.

See Also: Scotts Miracle-Gro platform gives you an IoT green thumb

Once all the smart devices are connected to the butler, it will begin learning patterns and start to automatically make decisions after seven days. If you boil the kettle at 8.30am each morning or turn off the lights at 11pm everyday, it will do that task for you.

Lutron and Amazon continue to add to smart home family

Lutron already provides a mobile app, available on iOS, Apple Watch and Android, for turning on and changing the light’s colors. The system also supports other smart home hubs like Works with Nest and Apple’s HomeKit, so this is a completion of the smart home package to make sure all consumers are happy.

These aren’t the first lights to take advantage of Amazon’s Alexa platform, Philips Hue, LIFX and Belkin WeMo are already integrated and offer similar functionality.

Lutron’s Caséta Wireless kit currently retails for under $100, and most wireless lightbulbs launched after 2014 are compatible with the system. You can check Lutron’s compatibility page for more information.

Hundreds of other devices are starting to integrate with Amazon’s platform, as customers start to realise the benefits of the Echo, Dot, and Tap in everyday life. Even platform hubs like the Nest thermostat and SmartThings are integrated with Alexa, taking advantage of the voice commands and APIs Amazon has on offer.

SmartAll supports over 1,000 devices now

It should be noted that while SmartAll supports over 1,000 devices, including Nest’s thermostat, the iRobot Roomba, and a selection of smart home door locks, it hasn’t launched any other smart home devices.

That might be a bit of a strain when it comes to connecting everything up and ensuring that devices work correctly. For one-function items like a door lock or a kettle that might not be too hard, but for a vacuum things like what rooms to clean and how long to clean may be hard to compute.

SmartAll already has an SDK ready for developers to build on the platform. It plans to add health management and emotion recognition to the hub in future updates, giving you food tips and changing songs to fit your current mood.

On Kickstarter, SmartAll is selling its butler for $199, though it has an early bird still available for the first 500, at $149. One year of cloud recording service costs an additional $100 or $299 with the butler.