The versatile humanoid robot Nao caught Gizmag's attention at the 2009 International Robot Exhibition (iREX 2009). What Nao lacks in size, he makes up for in features and capabilities. Nao can see (via two cameras), will react to touch, can surf the Web and can interact with other Naos. He can speak (in English or French, so far) by reading out any file stored locally in his storage space or captured from a website RSS flow. The bot is fitted with an accelerometer and gyrometer so he won't fall down, he's also equipped with two pairs of ultra-sound senders/receivers on his torso that give feedback on several echoes so Nao is aware of obstacles close by and can avoid them.

Nao’s vision is provided by two CMOS 640 x 480 cameras, which can capture up to 30 images per second. One is positioned on his forehead, aimed at Nao’s horizon, while the second camera is placed at mouth level to scan Nao’s immediate environment. And Nao’s software even lets you recover photos and video stream of his vision.

But Nao’s vision is more than just a movie-maker. His eyes can interpret his surroundings thanks to a set of algorithms in his on-board computer that can detect faces and shapes. This enables Nao to recognize the person talking to him, find a ball, and more complex objects. These algorithms have been specially developed for Nao by his makers, Aldebaran Robotics, and care has been taken to ensure they use up minimum processor resources.

Nao reacts to touch

Nao is fitted with a capacitive sensor on the top of his head that is divided into three sections which lets him react appropriately to touch. For example, pressing once can turn him off, or this sensor can be used as a series of buttons to register an associated action. The system comes with LED lights that indicating the type of contact that has been made. Nao’s designers say it is also possible to program complex sequences into him.

Nao can communicate in several ways. For local connections, infrared senders/receivers in his eyes allow him to connect to the objects in his environment, serving as a remote control. Yet Nao can also logon to your local network via Wi-Fi, making it easy to pilot and program him through a computer, or any other object that has a Wi-Fi connection. The Wi-Fi key is connected to the motherboard and accepts a, b and g standards.

Furthermore, Nao’s SDK lets you develop your own custom-built modules interfaced with OpenCV (the Open Source Computer Vision library initially developed by Intel). Nao’s makers say it’s easy to execute modules on Nao or transfer them to another PC connected to NAO because the OpenCV display functions allow developers to test their own algorithms with image feedback.

Nao can browse the Internet, of course, and interface with any website to send or retrieve data.

"Watching two Nao’s interacting was almost mesmerizing," reports Gizmag's Mike Hanlon. "At the show the two joked and played tricks and then performed a synchronized dance routine that kept most of the audience spellbound – not too difficult for robots with the same chipset and programming, but effective nevertheless."

"While the dancing was cute, it really showed off the possibilities of having the robots work in tandem to complete complex tasks such as geographic positioning or pooling analytical capacity."

No word on pricing but availability is getting closer.

View gallery - 4 images