Monthly Archives: September 2015

Raspberry Pi for the Talking Chatter Smartphone

You may have seen and even played with the Chatter Telephone, as the classic Fisher Price toy has been around since 1962. Although the design of the toy evolved over the years, in 2010, the Disney/Pixar film Toy Story 3 managed to depict the Chatter phone in its original classic form. Coinciding with the release of the movie, Fisher Price updated their model as the Talking Chatter Telephone. The new model carried some voice clips from the movie and had an all-plastic body, reminiscent of the boxy shape of the 1960’s original model.

Grant has used the famous Raspberry Pi or RBPi single board computer to turn the Talking Chatter Telephone into a Talking Chatter Smartphone. The modified phone gets its enhanced brains from an RBPi Model B+. All the logic comes from a Python script that runs on startup and access to the network is via a tiny Wi-Fi dongle. To retain the original factory look of the phone, Grant has avoided adding unnecessary buttons or screen and retained the original components and controls for interaction.

Internally, Grant has added a sensor to the hook cradle, but retained the original rotary dial that controls playback of the movie clips. For the eye movements, he has added a servo, while retaining the built-in speaker for providing audio output. He has retained the three micro switches the dial uses as a digital encoding system to sense the five dial positions. The RBPi replaces the battery box and the original PCB.

The phone pulls data from various online services via JSON format APIs. As of now, it connects to Forecast.io for receiving weather information and to Rotten Tomatoes for information on movies. However, one can easily extend this to other APIs as well. For example, the phone could be made to swap APIs for weather forecast with APIs providing current exchange rates, stocks or a Quote for the Day.

The phone also has a push notification system that uses a Twitter account. Grant uses it to pull new alerts from that account and to push them onto the account using IFTTT. He uses pre-defined alerts or recipes that he drives from his iPhone. For example, the IFTTT alerts the phone when Grant leaves work, there is rain forecast for tomorrow, the garden temperature is getting close to 10°C and when the International Space Station is about to pass overhead.

Grant has given the Chatter Smartphone a Wall-E startup sound for its boot up. That serves a dual purpose as the RBPi makes the sound when it has finished booting and lets Grant know that it is ready for an SSH connection. According to Grant, this should be made a standard feature for every RBPi. If you feel the volume of sound is low, it is easy to extend the RBPi’s audio jack to the back panel of the phone. This has the added potential of turning the Fisher Price phone into a high-end hi-fi system.

Using a Bluetooth add-on for the RBPi it is also possible to extend the audio to powered Bluetooth speakers outside the phone – making it wireless.

Using Raspberry Pi for Dog Training

Dog lovers and those who keep dogs as pets know that every dog has its own personality and like us humans, they too, often succumb to temptations. Some canines just cannot resist chewing those delicious new sandals, while others must investigate the leftover food scraps on the dining table. And, as working people cannot monitor their pet’s behavior the whole day, Dave Young took up a project involving lasers and the single board computer, the Raspberry Pi or RBPi, to help their pet keep its nose out of food scraps.

Dave Young’s Laser Dog Watcher consists of a laser tripwire, which silently alerts the RBPi as soon as the dog trips the invisible beam. The RBPi then takes a snap of the situation and plays an audio clip to dissuade the dog from its intentions. The project borrows the laser tripwire from an earlier design of a silent doorbell that Dave had installed at his home.

The laser tripwire system is of a simple reflection type. Both the laser source and its detector are in the same enclosure on one side of the room. On the other side is a small mirror placed to reflect the laser light back to the photo-resistive detector. Anything interrupting the beam also trips the detector, which sends out a signal. As ambient light plays an important role when detecting interruptions of the laser beam, Dave added a threshold level adjustment. In addition, to make it easier for the laser to hit the detector, he added a Fresnel lens in front of it.

The tripwire system is wirelessly linked to the RBPi. For this, Dave used an XBee module, series 1 by Digi, who has a starter kit for users. The integrated AD converter on the module transmits the detected light level digitally, which serves two purposes. Setting the threshold level for detection becomes simpler and there is no noise involved as during an analog transmission.

The XBee interface transmits the digital value to the receiver as a PWM signal. The XBee receiver filters this PWM signal to produce a DC voltage level. You need a second AD converter to translate the DC voltage level to a digital value. Although this method of converting the signal twice is somewhat cumbersome, the XBee module makes it very easy to create the wireless link.

XBee provides their XCTU setup utility, which helps in setting up the XBee radio transmitter and receiver. The starter kit also includes an XBee USB Adapter board that connects to the RBPi, which may need FTDI VCP or Virtual Com Port drivers.

When there is a break in the laser beam, the XBee radio alerts the RBPi, which then triggers its GPIO to play an audio file as well as takes a photo with the Pi camera. The audio file changes so that the dog does not get accustomed to the same reprimand each time it tries to break the rules. According to Dave, the camera is not essential, but it serves to detect false alarms and to record the surprised look on the dog’s face.

SOLI: The Final Interface is Your Hands

Finally, it is time to say good-bye to buttons and touchscreens. You need only wave your hands in thin air for controlling your gadgets. This game changer is a breakthrough from Google and is its project Soli. Soli makes it a thing of past to hit incorrect keys with your thumbs and you can conveniently forget swiping screens. The new gesture technology from Google is very precise and allows working on the smallest of displays.

Soli has small chips generating invisible radar to recognize finger movements. The chips are small enough so that they can be embedded into wearables and other devices. The deciphered finger movements are then translated into commands that computers can understand.

The system identifies delicate finger movements using the radar coming from tiny microchips. The system can use gestures to create touchpads, virtual dials and more as shown in the video above. Although there are camera-based sensors, such as Leap Motion, to capture gestures, they are cumbersome to set up requiring special hardware.

The inventor of this technology is Mr. Poupyrev, who heads the team of designers and developers at Google’s ATAP or Advanced Technology and Projects Lab in San Francisco. According to the Russian inventor, the beauty of Project Soli is the chip that can be embedded into just about anything and the use of invisible radar emanating from it.

Typically, police use the smallest radar for speed traps and even these are the size of a shoebox. The team had to struggle hard to shrink that radar to fit it inside a microchip. Although it took them 10 months, Mr. Poupyrev and his team were able to shrink all the components of radar down to millimeter size. They worked with Infineon, the German chipmaker and were inspired by the advances made in Wi-Gig, the next generation communication protocol for Wi-Fi.

Soli is a simple technology and the lack of cameras makes it easy to put wherever you want – in a toy, watch, wearable computer, car, furniture or anywhere. It is useful whenever people want to connect with devices. For example, Soli technology makes it possible to interact with objects in games making use of VR or virtual reality.

Since Soli makes it possible to replace a physical device, it works perfectly for Virtual Reality, as the field of vision of the user in VR is limited. The microchip uses radar to recognize movements of fingers. The chip radiates a broad beam radar for recognizing movement, distance and velocity. The radar uses the 60GHz spectrum and captures about 10,000 frames per second. The chip then translates the movements into commands that computers can understand.

Once Project Soli becomes reality, in the future we will be able to control devices such as fitness trackers and smart-watches only by our finger movements and will not require smart-phones as of now. Very soon, you may simply be able to snap your fingers to switch on the lights in your room and to vary its intensity by twirling your fingers.

An Action Camera for the Raspberry Pi

If you are the type that goes biking into the mountains and all the while recording your adventures on camera while on the trip, you need a camera that is biking-centric, robust and suitable for long-distance trips. Of course, several suitable cameras already exist such as the GoPro, Fly6 and the Sony Action CAM, but they are expensive not accessible to all. On the other hand, an action camera for the Raspberry Pi (RBPi) is not only cheap, it is also open-source and suitable for the purpose.

The design of the RBPi action camera is based on off-the-shelf components. It is very easy to build this project if you have access to a soldering iron and a 3D printer. Of all the models of the RBPi series, model A+ consumes the lowest amount of power, which is an important factor to consider since you will be running it on batteries.

If you are trying out the camera for the first time with an RBPi, using an open-source case is advised.

For an RBPi camera meant to be used for biking, three design goals must be met: the project must have a long battery life, be capable of wireless communication and its enclosure must be simple and made of durable material.

Apart from using the RBPi model A+, meeting the first requirement means using a large battery, especially if your rides are going to be multi-hour long. For the second requirement, it is necessary to have both Wi-Fi and Bluetooth, to make it easy to communicate with the camera. The last goal contributes to the first two, therefore, it must be given due consideration. Since the action camera is meant for outdoor use, making every port available outside the case would have reduced the structural integrity and its dust/water resistance.

To package everything into a small enclosure and ensure their working, you may need to work on the Wi-Fi dongle first, as that sticks out more than anything else does. For this, you may need to remove the USB jack and then remove the adapter from its plastic case. You can solder the wires directly to the board. The Bluetooth module may be placed on top of the RBPi and a ribbon cable used to connect it to the headers underneath. Next, make a support for the battery and its charger/booster so they fit snugly under the RBPi. You may need a few spacers to ensure the protruding headers do not puncture the battery.

Place the camera as close to the side of the RBPi and design the case to around all the components along with the RBPi. Usually, the case will be in two parts, with the camera module mounted on the top. Keep the camera module within the case and mount it in place with screws.

Use two buttons, with which the RBPi will start and stop the recording sessions. This may require you to use special scripts (you can use those by Alex Eames) for the RBPi to listen to a button press to start the camera and another button press to stop recording. Communication with the RBPi is done primarily through ssh.

Hear Only What You Wish To: Doppler ear buds

When there is a need for solitude, peace and quiet, some resort to earmuffs. Although good for cutting off or reducing the loudest of noises in your neighborhood, earmuffs cannot help you to hear the sounds of the world to your liking – either you hear nothing or you hear it all – there is no in-between.

Now Doppler Labs has an ear bud that allows you to choose how you would like to hear the sounds surrounding you – with a volume knob. The ear buds are no hearing aids and neither are they a pair of headphones. Once paired to the iPhone, you have a way to customize your hearing. You turn a volume control up or down until the sounds of the ever-louder world match your liking.

Wearing the Doppler ear buds can make your commute a little easier or make the concert sound as good in section 220 as in row 1. Doppler is presenting a long-term vision of its “hearables” technology that it expects everyone will eventually use in their ears, throughout the day.

According to Doppler, the world is becoming louder by the day. However, the ear buds control what you allow into your ears. You have a volume control and an equalizer for your ears. Control the loudness of the sounds you hear, crank up the bass or even mute sounds selectively, if you do not want to hear something – a baby crying, the screech of the subway, anything. As Fritz Lanman, the executive director of Doppler expresses, it is amazing what a volume control in your ear can do.

Doppler’s companion app on the iPhone has a dial graduated in decibels. Spin it in one direction to increase the loudness and reduce everything to a whisper by spinning it back the other way. The Effects section has buttons to help you choose the ambience. You can make sounds echo several times or you can choose a reverberation just as you were on stage in Carnegie Hall.

In fact, the Doppler companion app helps you to add many effects to sound entering your ears. There is provision for mixing different frequencies. You can make the songs flange, echo or add fuzz. The integrated noise-cancellation allows turning off most sounds – there is a special Baby Suppress button. Doppler has designed this morbid-sounding mode for muting the sound of crying babies.

This is not the first time people have attempted to augment hearing as something beyond hearing aids. A hundred years ago, the inventor of the first headphones, Nathaniel Baldwin made it as an amplifier and suppressor. Others have done considerable research on this subject.

Doppler has turned all this research into something you can wear. However, nobody likes to wear hearing aids, and Doppler thinks the key to making their ear buds acceptable to people is by making it absolutely clear what these things can and cannot do. According to Doppler, the ear buds are a niche product, mostly for music lovers, allowing them to tweak a concert to their liking. They are not for 24-hour wearing and not for making phone calls.

This Drone Avoids Obstacles When It Sees Them

If you thought drones could only fly and had to be manually guided around obstacles, the information you have is about five years old. Within the last few years, drones available to the average consumer have progressed by leaps and bounds. Most drones possess an onboard computer system that allows them to navigate autonomously. They can follow along with their owner or lead a path defined by GPS waypoints, capturing alluring aerial footages on the way.

Up until now, the drones that we came across were blind to their surroundings. They were able to capture photos, but if a ski lift or a big tree got in their path, the drones did not have the capability to change course to avoid it. With the First Guidance System from DJI, all that is now relegated to history.

The First Guidance System comes with a combination of stereo cameras and ultrasonic sensors. They allow the drone to detect objects as far away as 65 feet or 20 meters and take recourse to keeping itself at a preconfigured distance. This robust sense and avoid technology not only helps to integrate drones into everyday life, but also enables ambitious projects such as the Prime Air of Amazon.

Just as the robotic driverless car does, drones can now move about towns and cities, capturing new footages, delivering packages or even handing out parking tickets. According to DJI, research teams are using their guidance system for some unique applications. For example, Fudan University at Shanghai has created an aerial solution with Intel processors for aerial detection of illegally parked cars.

A new Matrice 100 drone from DJI powers the guidance system. DJI has made the system as a developer-friendly craft that users can modify for specific tasks across different industries, even acting as a testbed for experimental work. DJI is pushing this not only at the hardware manufacturers, but also as a platform for the entire drone industry.

On the drone, you will find additional expansion bays. These allow you to add components and customize the payload, allowing it to fly with any device of your choice. For example, you can put communication tools, computing boards, sensors, cameras and more into the sky. This allows you to complete your complex jobs from a birds-eye view, while the drone gathers data.

For example, using devices from DJI or third parties, you can connect and fly the drone and transmit data to ground in real time. With dual parallel CAN ports, the Matrice 100 connects DJI devices such as the Guidance sensor systems, while Dual parallel UART ports allow connecting third party components.

You can extend the flying time of your drone by up to 40 minutes with the help of an additional battery. The adjustable arm angle for each of the four arms allows greater yaw torque and response. The rigid, strong and lightweight carbon fiber frame of the Matrice 100 offers unmatched reliability and reduces stiffness. Soft vibration-absorbing material, lining the arms, eliminates nearly all feedback from the powerful motors. That keeps all critical components stable while allowing unparalleled accuracy.

Transparent Harvester of Solar Energy

Common belief is anything that harvests solar energy must be non-transparent. Popular logic is if sunlight is allowed to pass freely through the collector, it cannot lead to energy production. Although this may be partly true for the visible spectrum of light from the sun, it must also be considered that the sun gives out radiations beyond the band of light visible to the human eye.

Therefore, even see-through solar concentrators can successfully harvest energy from sunlight. Now, a team of Michigan State University researchers has proven this by creating a transparent solar concentrator. They claim to be able to turn any window into a photovoltaic solar cell. Not only windows, any sheet of glass, including the screen of a smartphone, can be turned into a harvester of solar energy. All the while, the panel remains truly transparent.

Earlier, transparent solar cells were restricted to tinted glass or compromised the visibility. This did not become popular, as people felt rather uncomfortable sitting behind colored glass making for colorful environments. In contrast, the new solar cell from the Michigan State University is completed transparent.

At MSU, researchers used TSLC or Transparent Luminescent Solar Concentrators. These employ organic salts for absorbing wavelengths of light normally invisible to the human eye, such as the infrared and the ultraviolet light. The researchers can tune the amount and composition of the organic salts to pick up only the near-infrared and the ultraviolet wavelengths leaving the visible spectrum untouched. The organic salts make the captured wavelengths glow at another wavelength – the infrared.

The TSLC then guides the infrared light to the edge of the panel, where it encounters thin strips of photovoltaic cells, which converts it to electricity. The organic salts do not absorb or emit any light in the visible spectrum and the panel looks extraordinarily transparent to the human eye.

The process is non-intrusive and opens doors to several opportunities of deploying solar energy creatively. Tall buildings with lots of windows can benefit tremendously with this technology, as can any mobile device demanding high aesthetic quality. The biggest benefit is you can have a solar harvesting surface and need not even know that it is present.

At present, the energy producing efficiency of TSLC is rather low, of the order of 1 percent, and additional work is needed to improve its performance. However, researchers are confident they will eventually increase the efficiency to above 5 percent. In comparison, non-transparent luminescent concentrators offer efficiencies of up to 7 percent.

In July 2014, the journal of Advanced Optical Materials carried an article describing the transparent solar cells. Apart from the lead researcher Richard Lunt, Yimu Zhao, Benjamin Levine and Garrett Meek are other members of the research team working on transparent solar cells at MSU.

Lunt has cofounded a Silicon Valley start-up – Ubiquitous Energy – for commercializing the transparent solar cell. The researchers have named the technology ClearView Power. They plan to integrate it directly on surfaces of mobiles, creating an auxiliary power source. They also want to promote this as a power-producing invisible coating for windows.

Zumo-George the Raspberry Pi Behavior Driven Robot

It is difficult to forget the roving Roomba, but it is time we have a new rover – Zumo-George. It is necessary to look differently at the process of control from a series of behaviors, while defining the tenets of development driven by behavior. Development undertaken via BDD or behavior-driven development is a superior method of emphasizing collaboration and communication between testers, developers and business stakeholders. Features and scenarios define behavior, as the Gherkin syntax specifically elicits –

Given: Zumo-George is more than 10cm from wall

Situation: Power is applied to the motors

Result: Zumo-George should drive forward

Developers write the BDD scenarios before writing other code, and this determines what code is written. This process reduces wastage. In addition, the written code drives the development, which, in most cases, passes the first time. As Zumo-George executes the scenarios, developers can see exactly what steps it passes, what it fails to pass and whether it encounters any situation that they have overlooked.

Intermixed with electronics, use of BBD to program robots such as Zumo-George can be an ideal abstraction for exploring robotic control based on behavior – BDR, or Behavior-Driven Robotics. Such programming can even include testing or internal diagnostics on Zumo-George. For example,

Given: Lights are all off

Situation: When light is switched

Result: light should turn

Or, on a lighter side,

Given: Batteries are fully charged

Situation: Shoot lasers

Result: Target should fry

As Zumo-George has no laser.

Zumo-George has to execute a series of internal diagnostic tests each time it boots up. If it fails any test, then it will simply refuse to rove and will flash a red light. This will preclude the problem of the robot running out of control.

Naming the robot Zumo-George, the developers prefer referring to the robot as a “he” rather than “it.” They expect Zumo-George to mimic certain human behavior. For example, do not bump into a wall while walking/driving.

Polulu’s Zumo and the Explorer HAT Pro from Pimroni form the basis of the rover (including its name). Therefore, Zumo comes in several variants. For example, for Arduino, there is the all singing Zumo 32U4, with accelerometers, LCD, buzzers, sensors and more. Then there is the bare-bones Zumo Chassis Kit and this is most suitable for Raspberry Pi (RBPi), as users can add their own electronics.

A Smart Fridge Tells You What It Wants

Imagine you are at the grocery store and wondering what you need for the next week – if you could only peep inside your fridge now, shopping could be easier. With the new smart fridge from General Electric in your kitchen, you could use a smartphone and ask the fridge what it lacks. The smart fridge will tell you exactly how much beer, soda, milk, and even how many eggs or separate vegetables it is left with. Actually, GE ran a contest taking ideas from users that could be turned into serviceable and manufacturable accessories. They announced the winners at the CES 2015 at Las Vegas. MakerBot Industries, LLC in Brooklyn, NY, is offering not only a MakerBot Replicator, but also a 3D Printer that allows engineers, traditional product designers and even consumers to prototype their ideas rapidly. Successful designs will be manufactured at FirstBuild at their microfactory in a fraction of the time it normally takes.

GE’s ChillHub is the first major home appliance that consumers can base their prototypes on to make their own accessories for a smart refrigerator. The ChillHub can tell your smartphone how much milk is currently left, because it has a milk weighing arrangement that your phone can query when you are at the grocery store. Besides the milk weighing arrangement, the ChillHub has several USB hubs allowing you to add your own plethora of smart accessories and sensors to let your smartphone see what else you need.

For proof of concept, GE and MakerBot, in collaboration with Thingiverse, came up with the Icebox Challenge, which had about 200 entries. The first-prize winning entry was an Odor-eating Hotshot. It uses a standard box of baking soda, but maximizes its odor-canceling capabilities, keeping track of its presence in the refrigerator and alerting users when to replace it.

The second prize was a bottle holder that helps the user organize different beverages while doubling as a chip-clip that keeps bagged snacks fresh. The third prize was the Butter Pig that dispenses standard butter sticks to simplify cooking in the kitchen, measuring of recipes and making toast.

The ChillHub is suitable for adding third-party accessories because of its eight USB ports. The ports are capable of delivering up to 2A each. That makes it very easy to add accessories that can be accessed from the Internet via their built-in Wi-Fi. GE calls the ChillHub architecture a community-generated product, which is based on an open-source iOS app. This app allows users to easily access the accessories plugged into the USB ports. Other fridge owners can hack their own appliance and make DIY upgrades using the FirstBuild.

FirstBuild community members conceived the design. They used 3D printers as a means of prototyping accessories quickly. The first accessory to be designed was the Milky Weigh that tells how much milk it holds. You can buy the complete Milky Weigh from FirstBuild, or if you are more adventurous, download the entire design and 3D print the components. The Green Bean circuit board from FirstBuild provides the electronics that actually weighs the milk for Milky Weigh.

A Slice of the Raspberry Pi

The Compute Module of the credit card sized popular single board computer, RBPi or the Raspberry Pi, is not an end-user product. Manufacturers can use the device when they require an ARM-based platform to build their devices on and sell. Therefore, computing hobbyists will find it difficult to get their hands on the Module if they want to evaluate it.

The RBPi itself is readily available to anyone who wants to buy and use it for projects. However, this Compute Module is not sold as such to hobbyists and for evaluating the Compute Module, it is necessary to get hold of a real product based upon it.

Five Ninjas, some people from the RBPi Foundation and the Pi-friendly accessories seller Pimoroni has a compact media player based on this Compute Module. Their product – Slice – was the result of inspiration based on the original Apple TV.

The first Apple TV was based on the x86 and was silver colored. This was eminently hackable, unlike the later iOS running black box that Apple made. People ripped out the custom Mac OS X installed, replacing it with a Linux desktop. They then added a more open, flexible media center, which ran XBMC.

The FiveNinjas Slice Media Player turned out to be more powerful than the modified x86 version of the Apple TV. The first few Slices have just left the Sheffield assembly plant of Pimoroni. Each has a custom motherboard with a single Compute Module in a DIMM-slot.

The Slice looks like a small metal box that has a translucent plastic spacer running all round the middle. The metal of the box is anodized aluminum in one of choice of three colors – red, gunmetal and black. The entire device feels and looks very stylish. Although you cannot see inside the box through the spacer, Slice puts out a very cool light through it. The light comes from Slice’s 25 NeoPixels. These are individually addressable RGB LEDs, with each containing an in-package controller.

The Slice uses these LEDs to create a rainbow of various color sequences. These sequences are triggered as the user interacts with the Slice using its remote control. While Apple had a slimline aluminum remote, Slice has a somewhat thicker one made of plastic.

Slice has 4GB of flash, which allows it to run any Operating System without a hard disk. It actually runs OpenElec, which is a simplified Linux distro capable of booting straight into Kodi, the media application. Therefore, users can simply play video and music files on their NAS or share from their computers.

Internally, Slice has a SATA connector mounted on the underside of the motherboard. Users can put in a small 2.5 inches disk drive and fasten it on to the motherboard within the case. There are four USB ports and users can hook up Slice to their computers to mount as an external drive automatically.

Currently, there is no app to control the display of colors from the LEDs. However, one is in development and will be available soon. The Compute Module uses a powerful 900MHz Broadcom SoC with a graphics core.