Performance comparison of leg detection methods using laser range finder

Human interaction becomes an important issue in the field of mobile robotics. To achieve humanfriendly naviga tion, the robot needs to recognize human on cluttered backgrounds, and this can be fulfilled by the detection of human legs. The detection of human legs is advantageous because it enables detecting environmental obstacles at such heights. In this pa per, we compared the performance of an algorithm using a single laser range finder（LRF） proposed in Ref. L 1 ] with that of wellknown feature extraction approaches bounding box and circle fitting proposed in Ref. [ 2 ] by using the same laser scanned image.

Micro Manipulation Using Magnetic Microrobots

When developing microscale robotic systems it is desired that they are capable of performing microscale tasks such as small scale manipulation and transport. In this paper, we demonstrate the transport of microscale objects using single or multiple microrobots in low Reynolds number fluidic environment. The microrobot is composed of a ‘U＇ shaped SU-8 body, coated on one side with nickel. Once the nickel coating is magnetized, the motion of the microrobots can be driven by external magnetic fields. To invoke different responses from two microrobots under a global magnetic field, two batches of microrobots were fabricated with different thicknesses of nickel coating as a way to promote heterogeneity within the microrobot population. The heterogeneity in magnetic content induces different spatial and temporal responses under the same control input, resulting in differences in movement speed. The nickel coated microstructure is manually controlled through a user interface developed using C＋＋. This paper presents a control strategy to navigate the microrobots by controlling the direction and strength of ex- ternally applied magnetic field, as well as orientation of the microrobots based on their polarity. In addition, multiple micro- robots are used to perform transport tasks.

Feedback Control of an Achiral Robotic Microswimmer

Magnetic microswimmers are useful for navigating and performing tasks at small scales. To demonstrate effective control over such microswimmers, we implemented feedback control of the three-bead achiral microswimmers in both simulation and experiment. The achiral microswimmers with the ability to swim in bulk fluid are controlled wirelessly using magnetic fields generated from electromagnetic coils. The achirality of the microswimmers introduces unknown handedness resulting in uncertainty in swimming direction. We use a combination of rotating and static magnetic fields generated from an approximate Helmholtz coil system to overcome such uncertainty. There are also movement uncertainties due to environmental factors such as unsteady flow conditions. A kinematic model based feedback controller was created based on data fitting of experimental data. However, the controller was unable to yield satisfactory performance due to uncertainties from environmental factors; i.e., the time to reach target pose under adverse flow condition is too long. Following the implementation of an integral controller to control the microswimmers＇ swimming velocity, the mieroswimmers were able to reach the target in roughly half the time. Through simulation and experiments, we show that the feedback control law can move an achiral microswimmer from any initial conditions to a target pose.