As an echolocating sensor moves through an environment the pattern of echoes reflected by objects to that sensor changes continuously, creating acoustic flow. Acoustic flow has been observed in both bats and humans. In this paper, we develop a theory of acoustic flow, and discuss measuring it with a Continuous Transmission Frequency Modulated (CTFM) ultrasonic sensor.

This paper describes the design and modeling of a robot that swings on the end of a tether. Its height above the ground is controlled by the length of the tether. Its location on a spherical shell at the end of the tether is controlled by a pair of ducted fans. The vertical orientation of its body is determined by the point at which the horizontal thrust force acts relative to the centre of gravity (cog). The system acts like a double pendulum with two modes of natural movement. It is modeled with coordinate frames at the ends of the tether and at the cog. Due to the second mode of motion it may oscillate around its centre of gravity while it swings or flies in a circle. From the dynamic model we develop a model for simulating the motion of the tethered robot.

The Draganflyer is a radio-controlled helicopter. It is powered by 4 rotors and is capable of motion in air in 6 degrees of freedom and of stable hovering. For flying it requires a high degree of skill, with the operator continually making small adjustments. In this paper, we do a theoretical analysis of the dynamics of the Draganflyer in order to develop a model of it from which we can develop a computer control system for stable hovering and indoor flight.

When a mobile robot is constructed the odometry must be calibrated. Calibrating a 4-wheel robot requires the accurate measurement of steering angle as well as translation and rotation. Some measurements are made with a tape measure and some with ultrasonic sensors. The measurements are used to determine the parameters of the odometry calibration matrix and the steering kinematic model. A procedure for calibrating a 4-wheel robot is discussed.

People have difficulty understanding ultrasonic sensing because they cannot see sound. The purpose of simulation is to overcome this problem by visualizing the scattering of ultrasonic waves off objects. The lattice gas model calculates wave behavior with finite difference equations to produce data suitable for grayscale visualization. This visualization is useful when designing ultrasonic sensing systems for navigating mobile robots. Situations that result in the sensor failing to detect an object can be studied with the simulator.

In the Titan project we applied a new approach to prototyping mobile robots by choosing tools which are commonly used by leading aerospace manufacturers and many other industries. We have gained substantial experience when using the LabVIEW real-time programming environment coupled with the industrial quality data acquisition cards, both are made by National Instruments. The methodology of virtual instruments software tools combined with the graphical programming environment was found to be very efficient for interactive cycles of design and testing, which are at the core of robotics prototyping.

The success of any mobile robot is dependant on its ability to perceive the environment in which it is working. This paper describes some work to verify the suitability of a new sensor as a landmark detector for a mobile robot. The landmarks that we are investigating are plants. An ultrasonic sensor continually transmits a frequency modulated signal. Echoes detected by a second transducer are demodulated with the transmitted signal to produce audio tones which are proportional to range. The spectrum of these tones is determined by the geometry of the object. An artificial neural network (ANN) is used to recognise plants. This paper discusses a series of experiments that prove that the machine perception system is independent of the range, size, and the orientation of the plant. This system has potential applications in industrial and office robots.

Discusses the development of a binaural ultrasonic sensor for mobile robot navigation, which is capable of measuring the range and bearing angle to multiple targets. A novel aspect of the design of the sensor is that it calculates the bearing angle twice, using inter-aural time difference (ITD) and inter-aural amplitude difference (IAD). These calculations of bearing are independent of atmospheric fluctuations and the geometry of the reflecting object. These independent bearing calculations are used to solve the correspondence problem. Echo pairs whose bearings differ by more than an acceptable amount are discarded.

Multiple reflection echoes are considered to be a source of errors in range readings because the simple model relating distance of flight to range is inaccurate. If the geometric path of the echo is known then it is possible to calculate the position of the reflecting object. In this paper, we developed a model of multiple reflections from the virtual source model. This model was validated by experiments with a mobile robot for the common situation of multiple reflection from a corner. By combining this multiple reflection model with the are model of ultrasonic sensing, we developed a virtual beam model which determines the path of a beam after multiple reflections. This model was validated experimentally by comparing its results to measurements made by a robot as it approached a corner. The models developed in this work are a first step toward solving the problems caused by multiple reflection paths for ultrasonic echoes.

Abstract With the upsurge of interest in parallel processing during the 1980s, some researchers have applied such techniques indiscriminately, even in applications where little inherent concurrency exists. In the paper the kinematics of a 5‐DOF robot arm are investigated to assess the degree of parallelism contained within alternate robot models. This concurrency is mapped into occam processes, then control software generated to control the robot arm, firstly from a single transputer, then secondly from a network of transputers. On the basis of our findings, we conclude that there is limited parallelism which can be exploited in the Homogeneous transform model of the kinematics of a serial‐link manipulator. An alternative formulation of the kinematic model is required before much speed‐up can be achieved by parallel calculations. However, the architectures studied are suitable for modular manipulator systems, where joint transforms have to be calculated separately.