This thesis examines the process by which an autonomous mobile robot constructs a map of its operating environment. This process can be considered as two distinct topics. First, the robot has to interpret the findings of its sensors so as to make accurate deductions about the state of its environment. This is the problem of ‘map-building’ Second, it has to select its viewpoints so that the sensory measurements contain new and useful information. This is the problem of ‘exploration’. This thesis describes a practical and experimental investigation into both of these issues.

This document is structured as a large number of short chapters. This reflects the wide range of subjects which had to be examined in order to build an effective working robot for map-building and exploration experiments. For ease of reading, the chapters are grouped into three parts; Part I (Chapters 2 to 4) examines the principal areas of previous research upon which this thesis is built; Part II (Chapters 5 to 10) describes the components of the map-building system; and finally Part III (Chapters 11 to 20) reports on experiments to evaluate the effectiveness of a range of exploration strategies. The closing chapters of Part III summarise the results and conclusions and suggest directions for further research.

The remaining sections of this introductory chapter serve as an overview of the thesis and put the later chapters into context.

The table on page 218 lists the input commands to ARNE's control software. The responses to each command are also listed.

ARNE's key physical component is a 300 mm diameter disc which supports the control electronics and the rotating sonar sensor. Below the disc is a chassis which holds the motors and shaft encoders to control the two drive wheels.

5.1 Hardware

ARNE has a drive wheel on each side of the chassis and a low-friction castor at the back. It moves holonomically, turning the wheels in the same direction to move forward or in opposite directions to rotate on the spot. Shaft encoders with a precision of 1024 steps per revolution determine the distance travelled by each wheel to a precision of 0.2 mm.

At the lowest level, the wheel movements are controlled by two dedicated HCTL-1100 motion control chips (Hewlett-Packard 1992, pages 1–77 to 1–115) which generate and execute trapezoidal velocity profiles. The length, acceleration and peak velocity of these movements are specified by the on-board CPU, a 68000-compatible ‘Mini-Module’ micro controller from PSI Systems Limited (PSI 1991).

ARNE's only range sensor is a single rotating Polaroid ultrasonic rangeflnder (Polaroid 1991) which can be seen in Figure 5.1 on top of the box which houses the CPU and other control electronics. The transducer is rotated by a stepper motor with a minimum step size of 1.8°. A full 360° scan is performed in twenty 18° steps.

Section 1.3 explained the decision to connect ARNE to a stationary workstation. A 9600–baud connection to the Mini Module's RS485 serial port was used for this purpose.

This chapter presents the sonar sensor model that was developed in this research. Figure 6.1 shows that the model is used to interpret the raw sonar returns from ARNE before the information is passed on to the the other modules on the workstation.

Section 6.1 outlines the operation of the Polaroid ultrasonic rangefinder used by ARNE. Section 6.2 then describes initial experiments to measure the range to a smooth wall in the test environment. The experiments highlight two key features of the sonar sensor: its wide beam and its uneven signal strength. Section 6.3 proposes a sonar model to mitigate the effect of these features by grouping neighbouring range readings. Section 6.4 then describes experiments to verify that the model will be applicable when measuring the range to the variety of objects that ARNE will encounter in the test environment. Section 6.5 then summarises the model.

6.1 The Polaroid Ultrasonic Sensor

Time-of-flight sonar is used in this thesis; distance information is derived from the time taken for a pulse of sound to travel to an object and be reflected back to the sensor.

Figure 6.2 is a simplified diagram of the rangefinder. Voltage pulses are sent to the transducer, which emits 16 cycles of square wave sound at about 50 kHz. As the sound begins, a timer is started. For a short period after transmission, the transducer is disabled (to give enough time for the vibration to die away) and it is then used to listen for an echo.

15.1 Motivation

This chapter describes a brief digression from autonomous exploration into human-guided exploration. The results in Chapter 14 showed Supervised Wall-Following to be an effective exploration strategy in environments with occlusion and traps. It was not, however, significantly better than simple wall-following in the ‘Empty’ environment. This raised the question:

Is it possible to improve the exploration performance in the ‘Empty’ environment or is Supervised Wall Following generating the best possible results, given the physical robot and its sensors?

To answer this question, experiments were performed to see whether a human operator, guided only by the developing map, could direct ARNE's movements so as to produce better results than Supervised Wall-Following. Similar experiments were performed in the more complicated ‘Walls’ environment.

15.2 Procedure

The exploration software includes an X-Windows interface which enables an operator to send commands (‘move forward’, ‘turn left’, ‘turn right’, and ‘scan’) directly to ARNE. This interface was used in the experiments described in this section. The interface also has the facility for the user to indicate, using the mouse, a position on the map to which ARNE should move. The system then plans and executes such a path. This facility was used for the longer movements between regions of interest.

Consideration was given to the choice of operator for these experiments. It was felt that a volunteer would have no experience of the way in which ARNE senses the world and builds the map and would therefore be unable to explore efficiently.

This Appendix gives the details of the calculations by which a line is fitted to a set of sonar observations.

It is initially necessary to distinguish two cases; the creation of a new confirmed line and the updating of an existing confirmed line. In both cases the confirmed line is fitted to a number of contact points, one for each sonar reading which corresponds to the line. The only difference is in the way the contact points are determined.

A confirmed line is created by ‘upgrading’ a cluster of elementary line segments. As explained in Section 7.1, each line segment has two contact points. Since segments are added to the cluster if they share a sonar reading with a segment already in the cluster, it is common for a single sonar reading to correspond to more than one segment. To avoid giving unnecessary weight to these ‘multiple’ readings, the confirmed line is fitted to a single contact point for each sonar reading. It is therefore necessary to calculate an ‘average contact point’ if the sonar reading corresponds to multiple segments.

If, on the other hand, the confirmed line is to be updated then the line already has a number of contact points and a new one is to be added. As explained in Section 7.3, a contact point is obtained by taking a point at the measured distance from the robot in a direction normal to the line.

This research places great emphasis on practical experimentation and quantitative evaluation of the results. To do this it is essential to have a precise measure of map quality. It is then possible to tune the map-building algorithms or to evaluate an exploration strategy by monitoring the quality of the map as exploration progresses. This chapter examines the issue of measuring the quality of a robot's map.

As an introduction, Section 10.1 examines some of the properties which one would expect to find in a useful quality metric, illustrating with examples of quality measures used by other researchers. The properties are that:

1. metric must be clearly defined.

2. metric must be multi-valued.

3. metric must reflect the purpose of the map.

4. metric must balance coverage and detail.

5. metric must be applicable during the construction of the map.

Section 10.2 discusses the need for an ‘omniscient’ observer. Is it possible for the robot to determine the quality of its map independently or can quality only be judged by comparison with a perfect map held by an external observer? It concludes that some quality measures can indeed by created by the robot independently but that measures derived from an ideal map are the most useful for the current purpose.

Section 10.3 surveys previous research in map-building, documenting the types of metrics which have been used. No ready-made metric was found which could be used in the current research.

This chapter describes the design and implementation of a localisation scheme for ARNE. Without such a technique ARNE's estimated position, based only on odometry, would gradually diverge from its true position.

The essence of localisation is to match recent sensory information against prior knowledge of the environment. Some researchers build a ‘local’ map from the latest sensor readings and then look for the best match between the local map and a global map. The correspondence can then determine the robot's position in the global co-ordinate system. Elfes (1989) does this by seeking the best correlation between local and global probabilistic grids. Crowley (1989) and Drumheller (1987) both extract line segments from the sensor data and compare the position, orientation and length of the each line with lines in the global model.

The experiments in Chapter 6 showed that it is impossible to determine either the type or position of environmental features from a single scan of ARNE's sensor. It is therefore not possible to construct a local map from each viewpoint. Instead, the technique adopted in this thesis is to find immediate correspondences between sensor readings and known features, and to use the range readings to known objects to estimate ARNE's position. Published examples of this approach include Curran (1993), Leonard and Durrant-Whyte (1992), Rencken (1993) and Kleeman (1989). The process of matching ARNE's sensor readings to known objects has already been described in Section 7.3.

ARNE's application requires it to follow efficient paths to user-specified delivery points. This chapter describes how these paths are planned.

Path planning serves two purposes in this thesis. First, it is obviously necessary to move during exploration and, although some of these movements (e.g. during wall-following) may be completely reactive and not use the map, others will require ARNE to go to a specified viewpoint while avoiding known obstacles. These movements will need to be planned.

Another, less obvious, need for path planning is in the measurement of map quality. As will be seen in Chapter 10, map quality is measured by predicting how successful ARNE would be at a number of test tasks, given the latest map. Path planning is needed to make this evaluation.

Section 7.4 described the construction of a free-space map from the list of confirmed features. Path planning is based totally on this map.

The planning technique used in this thesis was first presented by Jarvis and Byrne and is described by McKerrow (1991). A ‘distance transform’ is calculated which indicates, for any given cell in the free-space map, which of the neighbouring cells is closest to the goal. This information can be used repeatedly to generate a list of cells through which the robot can reach the goal.

Section 8.1 gives a brief overview of the technique and Section 8.2 gives the implementation details.

The paths derived from the distance transform are often unnecessarily jerky, giving a zigzag path to the goal.

Figure A.I is a simplified entity-relationship diagram which shows the main components of the feature-based map and the relationships between them.

Tables A.I to A. 11 list the data elements owned by each of the entities in Figure A.I. Note that all of the entities are contained, either directly or indirectly, within the “Map” entity. This reflects the fact that the map is implemented as a single shared data structure within ‘C’.

This book is the product of my PhD research at University College London. I am grateful to the many people and organisations that have made my research both possible and enjoyable. I have benefited greatly from the companionship and support of all of my colleagues in the Computer Science and Anatomy departments during the course of this work. The following paragraphs can only recognise some of the more direct contributions.

Many thanks are due to Michael Recce for his enthusiastic and constructive supervision. Michael has been generous with his time, his ideas, and his lab space. I am also indebted to Michael for his careful reading of draft versions of this document and his valuable suggestions about its style and content.

Jim Donnett has made many much-appreciated contributions to this work, ranging from hardware design and debugging through to a detailed reading of the thesis. Jim's breadth and depth of knowledge have been invaluable, and his friendship and sense of humour have helped me through some trying moments.

I owe a great deal to Clive Parker for the construction of my robot, ARNE. Thanks to Clive's electronic and mechanical skills, a loose collection of components was transformed into an effective research tool.

David Brown of the Statistics Department kindly took the time to advise me about the statistical analysis of my results, despite having been ‘Volunteered’ for the job. His insight and suggestions were most welcome.

17.1 Motivation

The exploration strategies presented so far in Part III of this thesis have differed in the extent to which the map has been used to control the navigational choices. Wall-Following (Chapter 12) and Longest Lines (Chapter 16) were both totally reactive, not using the map at all. Supervised Wall-Following (Chapter 14) used the map to detect circumstances in which wall following was becoming ineffective. Chapter 15 showed the results that could be obtained when a human operator used the map to guide the exploration. This chapter will present an exploration strategy in which ARNE's decisions are driven primarily by the information present in the partially-formed map.

The implementation described in Section 17.2 builds on the ideas presented in Section 4.2; ARNE approaches the interesting regions of the environment. The central issue is, of course, the definition of ‘interesting’. The definition adopted here focuses on the edges of free space, the regions in which free cells are next to unknown cells.

Section 17.3 presents the results of experiments to evaluate this strategy and Section 17.4 summarises the results.

17.2 Implementation

The first step in this implementation was to identify the cells on the free-space map which were to be examined. The decision was made that ARNE should approach unknown regions but, to avoid collisions or panic stops, it should not actually enter unknown regions. The interesting cells are therefore those on the boundary between free and unknown space.

The experience gained during the development of this thesis has suggested a number of directions in which the research could be extended. This final chapter examines these ideas under four groupings:

1. • Mixing planning and reactive navigation.

2. • Modifying the exploration method as the exploration progresses.

3. • Testing new sensors and new environments.

4. • Examining the feature map for inconsistencies.

A section is devoted to each of these areas.

20.1 Mixing Planning and Reactive Navigation

The Supervised Wall-Following strategy has shown that effective exploration can arise from a combination of reactive and model-based decisions. The application of the quality metric to maps of the ‘Walls’ environment showed that small errors in the map could lead to collisions unless the robot's movements took into account the latest information from the robot's sensors. These results suggest that it would be useful to extend the current research by implementing a navigation strategy which combines planning and reactive components.

There are clear parallels between this idea and the concept of compliance in automated assembly (McKerrow 1991, page 293). In both cases the robot uses its stored understanding of the state of the world to plan its actions, but it has to adjust its behaviour if sensory input disagrees with that understanding.

The work of Pay ton, Rosenblatt, and Keirsey (1991) is attractive in this context. They propose the use of ‘internalized plans’ which act as information resources to guide the reactive behaviour of the robot.

This thesis has described an investigation into the complementary problems of map-building and exploration by a mobile robot. This chapter highlights the most significant results of this investigation.

The novel contribution of this research can be summarised as:

• • The integration of a physical robot, a sonar model, map construction algorithms, and a localisation algorithm into an effective working system;

• • The definition and implementation of a novel quantitative measure of map quality;

• • A thorough quantitative and statistical evaluation of the map-building and exploration capabilities of the system, using the quality metric and a variety of exploration strategies.

The system components and the quality metric were described in Part II of this thesis. Sections 19.1 to 19.4 briefly review these topics. The experimental evaluation of exploration strategies formed the bulk of Part III of this thesis. The results of this work have already been summarised in Chapter 18.

Chapter 2 described the continuing debate between the ‘traditional’ supporters of modelbased robotics and the proponents of behaviour-based robotics. An outcome of the current research has been an awareness of the need to balance these two approaches. The value of reactive navigation became more apparent as the research progressed. Section 19.5 reviews the course of the research in the context of the ‘models versus behaviours’ debate.

19.1 The Ultrasonic Sensor Model

Chapter 6 presented a set of experimental results which showed how the Polaroid ultrasonic rangefmder detected each type of object that would be encountered in the test environments.

Chapter 12 showed the loss of map quality which arises as odometry errors accumulate and ARNE's position estimate becomes increasingly inaccurate. Chapter 9 presented a method by which ARNE can improve its position estimate by measuring the distance to known objects in its environment. The following sections describe the implementation of this localisation method and show the results of experiments to test its effectiveness.

A key component of the localisation method, the plant model, models the growth in positional uncertainty as ARNE moves. The plant model requires parameter values which are specific to the individual robot. Section 13.1 describes experiments to check that the parameters were approximately right for ARNE.

Section 13.2 then repeats the experiments from Chapter 12, but this time with the localization system in place, and compares the results. The loss of quality in the later stages of exploration is eliminated.

After the benefits of localisation have been demonstrated, Section 13.3 presents the results of wall-following with localisation in two other, more complicated, environments. The quality is shown to increase more slowly and to reach a lower maximum value in more cluttered environments. The reasons for this loss of quality are discussed.

The results of wall-following are then used to determine the best value for one of the central parameters of the map construction process, the confirmation threshold. Section 13.4 describes the experimental basis on which this choice is made.

Chapter 2 described numerous maps which have been used by mobile robots. This chapter considers which type of map to use in the current research.

The choice of map type is strongly constrained by the proposed application of the robot. In Chapter 1 a delivery application was chosen. Section 3.1 describes such an application in detail.

Section 3.2 uses the knowledge of the application to choose the maps to be used in this thesis. One of the most important choices was between probabilistic grid-based maps and feature-based maps. Section 3.3 explains why feature-based maps were selected. The chapter concludes in Section 3.4 by explaining why the robot will build its own map, instead of being given one by its operator.

The details of the map construction algorithm can not be described without knowledge of the robot and its sensors. This description is therefore postponed until Chapter 7 to follow the descriptions of the hardware and the sensor model in Chapters 5 and 6.

3.1 The Application

The choice of world model is strongly influenced by the proposed application of the robot. Indeed, as was discussed in Chapter 2, some applications do not require a world model at all. It is therefore vital to be precise about the intended application of one's robot before designing the world model.

This thesis addresses the construction of maps for use in an application with the following features:

Chapters 12 to 17 will examine individual exploration strategies and compare their results. This chapter introduces this part of the thesis by examining some general issues which are important whichever strategy is being tested.

Figure 11.1 shows the possible links between exploration algorithms and the rest of the system software. Two extreme types of exploration are represented on the diagram. The top portion of the ‘Explore’ box depicts reactive exploration in which the movement commands are based solely on the most recent sonar readings and the result of the previous command. In contrast, the lower portion depicts exploration which is totally map-driven. The experiments described in this part of the thesis investigate the potential benefits of striking a balance between these two extremes.

What would it mean to say that one exploration strategy is better than another? A reasonable interpretation would be that the first strategy produced a higher quality map than the second, for the same cost of exploration. One then has to decide how to measure the ‘cost of exploration’. Section 11.1 considers some alternatives and selects ‘the total time taken by the robot's movement and sensing actions’.

To make a fair comparison, the strategies must be tested in a variety of circumstances. The effectiveness of a strategy can depend on the environment being explored and on the starting position of the robot within that environment. Each of ARNE's strategies is therefore tested in at least 3 different environments and from 10 starting positions spread throughout each environment.