Energid’s Actin is used for control of robotic satellites, design validation of NASA rovers, and control of NASA’s Robonaut
Control and validation of on-orbit robotics for DARPA Phoenix
Energid’s advanced software enables robotic repurposing and reuse of retired satellites.
The DARPA Phoenix program was a multi-year program designed to develop and demonstrate technologies to cooperatively harvest and re-use valuable components from retired, nonworking satellites in geosynchronous orbit.
The Phoenix vision was to field a robotic satellite servicing spacecraft. The spacecraft employs grasping arms that manipulate unique tools to reconfigure orbiting satellite hardware. The arms are designed to have many degrees of freedom and the tools, newly developed under the program, will be innovative. This advanced robotic system will be controlled using Energid’s software.
From having an articulated arm in orbit to repurposing an abandoned satellite to finding new ways to launch, Phoenix is targeting many firsts.
Energid’s Actin software for robot control and simulation was ported to a real-time operating system (RTOS), and a set of libraries was created for real-time robot control on the servicing spacecraft. This control used the extra degrees of freedom in the manipulating robot arms to optimize for strength, avoid joint limits, and avoid collisions.
Energid’s software supported validation of the system through simulation during development and support training in preparation for launch. Actin enables automatic construction of control systems and simulations from CAD models, such as would be made with SolidWorks. This capability seeks to allow robot designers to test concepts in just minutes and accelerate the design process for Phoenix.
Design Validation of NASA Rovers
Rapid design iteration yields faster time to flight and reduced development costs
Energid has created a comprehensive software infrastructure for rapid validation of robotic designs. The software supports push-button validation through existing commercial design software, such as SolidWorks. After validation is invoked through the design-tool GUI, the robot design transfers to Energid’s Actin for kinematic and dynamic analysis. Actin allows interactive or programmatic placement of any number of end effectors on any number of mobile or fixed-base mechanisms, each with any number of kinematic links and branches.
Energid’s validation software provides dynamic simulation, including articulated-body and impact dynamics. The software allows link descriptions, end-effector descriptions, control algorithms, and the environment to be arbitrarily exchangeable as modules. Extensible Markup Language (XML) is used for configuration, data transfer, and exchange.
This software was applied by NASA Ames Research Center to facilitate field tests with a digital simulation of their K10 rovers. This new proxy simulation serves as a replacement for robot communications, actuation, control systems, power systems, sensors, environmental interactions, and behavior. It incorporates physics-based modeling of wheel-terrain interaction and obstacle collisions. The software is generic and applicable to all of NASA’s prominent mobile robotic vehicles.
Controlling Kinematically Redundant, Bifurcating Manipulators
Through a project with NASA, Johnson Space Center, Energid developed algorithms and software for controlling kinematically redundant, bifurcating manipulators. Our unique specialty is that our control algorithms can be applied to virtually any type of mechanism. From simple three-link mechanisms to complex 50-link robotic devices, control algorithms and manipulators are configurable and network exchangeable through a common interface using Extensible Markup Language (XML).
The control algorithms support multiple cooperating manipulators and exchangeable, flexible end-effectors. The topology of each manipulator is represented using a link tree, where any link can have any number of child links. The kinematics of each link’s joint can be represented using generic table functions. The physical extent of each link is described using a polygon mesh, with each polygon having a unique XML-describable surface property. Actuator properties, including motor inertia and friction, as well as dynamic joint controllers, are part of the system dynamics.