Energid Actin 5 simplifies real-time adaptive robotic motion control

Cambridge, Mass.-based Energid introduced at IMTS 2018 Actin 5, the newest version of its real-time adaptive motion control software. The new Actin 5 features are designed to further reduce the go-to-market time for a robotics developer looking to bring an application to market.

The Actin software, which figures out a robot’s most efficient path from Point A to Point B while avoiding collisions, now has a robot controller development framework that is configurable through XML for application on a range of platforms, including real-time operating systems. The framework’s modular design and XML configuration allows developers to embed Actin’s kinematic control into existing robotic systems.

“One of the hardest parts of building a robot arm is moving all the joints to get a particular motion. At the end of the day, everybody cares about how the end-of-arm tooling moves,” says Energid CEO Neil Tardella. “What you don’t really care about is how all those joints move to enable that motion of the end-of-arm tooling.”

Tardella tells The Robot Report Actin 5, which is now available, makes it easier for developers who are not well-versed in advanced programming languages. Energid added to Actin 5 an EcScript scripting language that can create motion commands with a few lines of code. Tardella says developers can still use C++ for advanced features, if desired, using the Actin 5 SDK’s documentation for C++ APIs.

Finally, a human-machine interface toolkit in Actin 5 allows developers to create touchscreen- or teach-pendant-friendly UIs. The HMI toolkit is built on Qt QML for multi-platform UI development and provides access to Actin features through QML properties.

Actin 5 demos at IMTS

Energid, which was acquired by Teradyne earlier in 2018, is demonstrating Actin 5 at IMTS, using a bin-picking application and external tool center point (TCP) application, both of which use cobot arms from Universal Robots (UR), also owned by Teradyne.

The bin-picking demo, which was developed as a URCap, uses shiny metal objects in a deep bin that are difficult to grasp. The system uses a 2D/3D image and custom pneumatic gripper developed by Energid.

“With this robot, we just taught the UR where the bin is and what to avoid. Actin runs everything in the background, figuring out how to best get from Point A to Point B,” says Tardella. “A lot of bin-picking applications focus on picking and not placing. So we’re trying to show placing as well. Actin knows how its picking up the part. Motion planning to placement is very important for certain applications.”

When it comes to bin picking, Tardella says a lot of emphasis, rightfully so, has been placed on perception. But he says without getting the motion control piece right, you won’t have a broadly adopted solution.

“A lot of systems only have one orientation at which they’ve been trained to pick up objects, and that can be limiting,” Tardella says. “Actin can take advantage of the pneumatic gripper that is capable of picking up products at various angles. Things like multiple pick points on a single object, that’s a motion control issue. Some systems struggle to pick up objects in deep bins because they don’t have the collision avoidance to model the bin and get around it. Some systems have long end effectors, but they don’t actually go into the bin.”

Nexcom, located next to Energid, is using one of its 7-axis robot arms for an Actin demo that showcases advanced kinematic solving, joint limit avoidance, and collision avoidance for complex applications. “A 7-axis robot is one thing, but you can imagine a day where you have robots with 9 or 10 axes or a two-armed robot with 14 axes,” Tardella says. “The process is exactly the same with Actin – it doesn’t care how many axes you have.”

Schunk’s IMTS booth is showing an Actin demo, with object tracking, of a Schunk SVH gripper and UR5. The system adapts dynamically to handle changes in the environment without the need to reprogram. The Actin software enables coordinated control of the hand and arm so the movements seem more human-like. The video below replicates the in-booth demo:

“We see a day where Actin is a component of a motion control stack,” says Tardella. “You have the real-time operating system, low-level drive control, inverse kinematics, and it should be something people don’t need to worry about.”