In general, pneumatic actuators have a good force-to-weight ratio making them almost ideal for biped walkers. However, it is hard to control the velocity or motion between the fully retracted and extended states. Stepper motors are highly controllable, but require large gearheads to produce the necessary power to actuate an ankle. These gearsheads often reduces the force-to-weight ratio to an undesirable level for a design.
The Hybrid Parallel Actuator (HPA)is being developed as a power source for the ankle of the JayWalker testbed. It consists of a pneumatic actuator and a stepper motor attached in parallel to an output shaft. The pneumatic actuator acts as an amplifier to boost the force required to move the ankle into the range of the stepper motor. The stepper motor is then used to control the actuator. This allows the HPA to have an improve force-to-weight ratio than the gearhead and improved controllability of a pneumatic actuator.
Biped robots offer advantages as personal service robots and terrain exploration robots. These advantages include navigating human environments better than wheeled robots, and needing only discrete footholds to move across rough terrain unlike the continuous paths required by wheeled robots. The mobility of biped robots has greatly improved over the years, they still have difficulty moving across rough and unstructured terrain. It is the goal of ISAL to help solve these problems by developing new rough terrain control schemes and mechanisms.
The JayWalker Testbed is a 3-legged, 2-D biped walking robot. It is modeled after the McGeer passive walker, except all the joints are actively controlled. The purpose of this testbed is to provide a platform for testing rough terrain walking strategies for biped robots.
Autonomous robot recovery is an important area of research for mobile robots. As mobile robots perform more tasks in hazardous environments. It is important to have a way to recover these robots without sending people into the hazardous environment. The basic idea of autonomous recovery is that as the mobile robot records its navigation path as it moves through an environment and if it looses connection with the base control unit, the robot will go into an autonomous mode and follow its stored path back to the base station until connection can be reestablished.
A key subsystem for autonomous mobile robot navigation is the ability of the robot to self locate in an environment. The ISAL has developed an algorithm which matches rangefinder data points to map lines in a quick and robust manner avoiding complicated feature recognition. A laser-based range data collection system was developed for a mobile robot based on this self-localization algorithm.
Telerobots are remote controlled robots that are programmed to respond to commands from a human operator, or to mimic the movements of its human operator. These systems offer an alternative to exposing humans to such hazardous work as maintenance of nuclear power plants, high voltage transmission lines or performing undersea maintenance of oil platforms.
Shared control involves development of sensors, software and speech recognition techniques that can supply some of the control input, and thereby increase the system's capability by enabling the telerobot to do more jobs autonomously. The ISAL has developed a shared hybrid position-force control system for electrohydraulic manipulators.
The realization of automation for a robot-tended machining workstation for a machining cycle (loading, fixturing, machining, unfixturing and unloading) depends partly on integrated control and automated process planning.
The workstation software, a major part of automated fixture planning, was developed by the ISAL. It consists of a set of controller modules and a robot program generator. The workstation controller centrally schedules robot-tended machining workstation activities which include the robot-tended fixture setups and takedowns, the fixture clamping actions, the machine tool program executions and file transfers. The robot program generator outputs robot instructions for the robot operation as a part of the process planning system.
