Children experiment with roBlocks

LEGO is perhaps the most universal childhood activity; generations of children have been entertained by the little plastic blocks that can be put together to form absolutely anything. However, the all-time favorite toy now has a high-tech competitor created by none other than Carnegie Mellon specialists.

Designed by graduate student Eric Schweikardt and architecture professor Mark Gross, “roBlocks” are small, magnetic blocks with computerized circuits inside them that can be put together to form robots. The blocks were designed as part of a project for developing computationally enhanced building sets.

“We’re looking at ways of embedding computational technologies into toys and construction kits to provide children opportunities to design things,” Gross said.

The roBlocks fit well with this central theme. “Everybody likes LEGO, so [we wanted to know] what happens if you put tiny little computers into all of the LEGOs,” Schweikardt said.

Following this idea, the team created an initial set of blocks by hand. Each block has a specific function and can be connected magnetically to other blocks. The way the blocks are connected determines the function of the robot that is finally formed. Children can thus combine the blocks in multiple ways, creating a variety of different robots.

There are four main types of blocks, Schweikardt said. Sensor blocks connect the robot to the external environment. They have infrared sensors to sense distance, light sensors, sound sensors, and knobs which can be manipulated by hand to vary the amount of signals received by the block. Actuator blocks use the data from the input blocks and produce the output. One type of actuator blocks has LED bar graphs to display the amount of input.

However, as Schweikardt said, “The motorized blocks are probably the most fun.” Motorized blocks have treads, which enable the blocks to move around. The treads act like wheels and roll to propel the robot forwards.

The third type of blocks are operator blocks. These manipulate the data, which is inputted into the sensor blocks. The blocks can invert values, add values, or choose the minimum of two values. The last type are utility blocks, which are basically batteries that run the entire combination. The results that can be obtained by combining all these blocks are fascinating. As demonstrated by Schweikardt, a robot that moves away from a rapidly approaching source can easily be built using these blocks. The team however, does not want to cease experimenting at this point.

“We’re building as many blocks as we can to try and build bigger robots,” Schweikardt said.

Blocks that are still in the pipeline include actuator blocks with hinges and blocks with rotators. The blocks are mainly designed to be an educational tool for children in the age group of 8 to 12 years.

Schweikardt said that children who were around 8 years old could start playing with the blocks and make simple robots with them. Older children would start understanding how the blocks worked together to form the entire robot.

The simple logic behind the functions of the robot should help the children enhance their lateral thinking skills. According to a Carnegie Mellon press release, the blocks can also be reprogrammed to perform slightly different tasks from what they were originally designed to do. For example, the five logic blocks known as “and,” “or,” “not,” “nand,” and “xor,” enable children to improve how their robots react to different stimuli.

The names of the blocks are based on logic gates that carry out logical operations on inputs and form a unique logic output. These logic gates are usually an integral part of digital circuits. While the blocks are valuable educational tools, they serve a higher purpose as well.

“People have a hard time thinking about complex problems,” Schweikardt said. “The problems are more complex, meaning that there are a lot more individual actors or players, all obeying different rules. You don’t know what sort of change [in one part] will affect everything else.”

These complex problems can be solved by examining patterns in the roBlocks. Unlike contemporary robots in which a single “brain” controls the entire robot, robots formed with roBlocks are made up of individual parts that have different functions, yet work in unison to form the entire model.

“This [complex robot made of smaller blocks] is like a model system where each little robot is its own actor. This is like taking six or eight or 12 or hundreds of little brains and putting them all together and seeing what happens,” Schweikardt said. By studying such a model, people can understand how different elements of a system can work together to perform work efficiently and reduce problems.

Schweikardt and Gross hold high expectations for the blocks. “We expect people to really like them and want them and we might even go to mass marketing,” Gross said.

Initially, the group wants to popularize the blocks through museums and science centers. “We’re working with a company that will bring these blocks out first for use in museum settings, after-school settings, and informal learning settings,” Gross said.

As stated in the Carnegie Mellon press release, Schweikardt and Gross expect to have the blocks in the market by December 2008.