We’re starting to see swarms of heterogeneous robots, with different robots combining their powers to make each other more efficient and more capable. One of the first projects to really make this work was Swarmanoid, with teams of footbots and handbots and eyebots, and researchers presented a similar idea at IROS earlier this month, using an augmented-reality drone to help a swarm of self-assembling ground robots… read more

In this video, the Pet-Proto, a predecessor to DARPA’s Atlas robot, is confronted with obstacles similar to those robots might face in the DARPA Robotics Challenge (DRC). To maneuver over and around the obstacles, the robot exercises capabilities including autonomous decision-making, dismounted mobility and dexterity.

The DARPA Robotics Challenge will test these and other capabilities in a series of tasks that will simulate conditions in a dangerous,… read more

A new robot developed by David Dorhout and colleagues from Dorhout R&D is designed to plant seeds in a field while coordinating with a gang of other robotic farmhands, according to New Scientist.

The robot can walk in any direction while avoiding obstacles, using a sensor underneath its body to detect where seeds have already been planted. Once it finds an untouched patch, it drills a hole… read more

Microsoft Research | Deep Neural Networks for Speech and Image Processing: Neural networks are experiencing a renaissance, thanks to a new mathematical formulation, known as restricted Boltzmann machines, and the availability of powerful GPUs and increased processing power.

Unlike past neural networks, these new ones can have many layers and thus are called “deep neural networks”; and because they are a machine-learning technique, the technology is also known as… read more