Autonomous Systems: Spider Workers

“SiSpis are part of a larger picture that we define as Siemens Agile Manufacturing Systems (SiAMS) and they represent the core of our autonomous systems research here in Princeton,” says Dalloro, who explains that the spider-like bots are essentially fully autonomous additive manufacturing devices with legs.  “We are looking at using multiple autonomous robots for collaborative additive manufacturing of structures, such as car bodies, the hulls of ships and airplane fuselages.” Additive manufacturing, also known as 3D printing, is a technology in which products or surfaces are created or refined by depositing layer after micro-layer of a selected material.

Of course, adding a layer of material to the inner surface of something the size of a ship’s hull is a job that would require more than just a few spiders. Most likely, hundreds would be needed. So the key question is: how would such an army of robots work together? The answer is a form of autonomy. “Each spider is capable of manufacturing only a small portion of a work piece,” explains Hasan Sinan Bank, who has played a leading role in the project and has filed multiple associated patents. “We are therefore trying to conceptualize and optimize the kinds of collaboration these robots should engage in.” This process is supported by algorithms developed by Dalloro’s team that allow multi-robot task planning so that two or more devices can collaborate on the additive manufacture or surface processing of a single object or area.

To accomplish this, the robots use onboard cameras as well as a laser scanner to interpret their immediate environment. Knowing the range of its 3D-printer arm, each robot autonomously works out which part of an area – regardless of whether the area is flat or curved – it can cover, while other robots use the same technique to cover adjacent areas. By dividing each area into vertical boxes, the robots can work collaboratively to cover even complex geometries in such a way that no box is missed. “No one else has attempted to do this using mobile manufacturing,” says Bank.

Autonomy is also a central part of other spider behaviors. For instance, the spiders know where they are in a given environment. When its batteries are low after about two hours, a spider will find its way back to a charging station, but not without first transmitting its data to another spider that has just been charged, thus allowing the second spider to pick up exactly where the first one left off. Also, whenever they encounter an obstacle, the spiders autonomously find a path around it. Siemens’ spider-worker project was initiated in January 2014. “Its goal,” says Dalloro, “was to create a prototype platform for autonomous manufacturing machines that can understand a task, split it up among available robots, and enter into a manufacturing process in a collaborative and coordinated way without explicit programming.”

Thanks to NX, a Siemens PLM software solution, and a hybrid software developed by Dalloro’s team that combines NX with ROS (Robot Operating System), the spiders quickly began taking shape. “With the exception of the spiders’ mini motors and cables, which were off-the-shelf products, we developed everything ourselves from the mechanics to the software,” says Dalloro. True to the project’s focus on advanced manufacturing, each time a spider component was designed in the virtual world, it was produced using 3D printing.

In addition, the team had to design the robots’ behavior. “To do that,” says Bank, “we had to develop software tools to simulate their behavior across communities of robots.” That meant that the team had to come up with a method to precisely calibrate the spiders’ manufacturing nozzles. Each spider is equipped with an extruder similar to those on traditional 3D printers, but is limited – for now – to using a cornstarch-and-sugarcane printing substance known as poly lactic acid.

Where do the spiders go from here? Clearly, the researchers have achieved their initial goal of developing a system that is characterized by maximum autonomy and minimum programming requirements. “Once the technology becomes mature,” says Bank, “It could be applied to almost anything.”

From Siemens: See more here