RoBIOS/project.html

   1 Date: Fri, 21 Jan 2000 17:08:17 +0800
   2 From: Thomas Braunl <braunl@ee.uwa.edu.au>
   3 To: pere@ee.uwa.edu.au
   4
   5 Hi Petter
   6
   7 here is the text passage from Klaus and I put in the headlines for
   8 the requested sections.
   9
  10 Since we want funding for 3 Ph.D.s I suggest the project structure:
  11   - robot operating system
  12   - robot simulator
  13   - behaviour-based application
  14
  15 What do you think?
  16 Thanks
  17   Thomas
  18
  19 P.S. can you handle MS Word files as well?
  20 ________________________________________________________________________
  21 Thomas Bräunl                                e-mail braunl@ee.uwa.edu.au
  22 Dept. of Electrical & Electronic Engineering       phone +61 8 9380-1763
  23 The University of Western Australia                fax   +61 8 9380-1168
  24 Nedlands, Perth, WA 6907, Australia     http://www.ee.uwa.edu.au/~braunl
  25 ________________________________________________________________________
  26
  27
  28 17.1 Aims, significance and expected outcomes
  29
  30 Aims
  31
  32 In the broad field of robotics and especially in the field of the
  33 Autonomous Mobile Robots (AMR) the subject of service robotics has
  34 received growing attention in recent years. Service robots need to be
  35 intelligent and be easy to handle by the growing number of
  36 non-professional operators. To fulfil this demanding task, alternative
  37 solutions to the traditional approaches have to be found.
  38
  39 The classic approaches use the well-known Artificial Intelligence (AI)
  40 planning methods to solve a given problem [LATOMBE 91]. They can only
  41 cope with situations which the programmer has foreseen and implemented
  42 in the robot's control logic. The robot will be helpless if something
  43 unpredictable happens so that the pre-calculated actions cannot be
  44 performed.
  45
  46 The next step in this development was the biologically inspired
  47 behaviour-based approach [BROOKS 86]. It is based upon a set of simple
  48 behaviours which are independent of each other. Each individual robot
  49 behaviour consists of coupling sensor inputs to actuator outputs to
  50 produce a desired action. The behaviours are organised in priority
  51 order. A so called arbiter chooses the action with the highest
  52 priority and therefore controls the total behaviour of the robot which
  53 makes it appear to be intelligent. This approach has proven to be
  54 robust in the real world with minimal environment modeling. However,
  55 it has a limited capability to learn or acquire knowledge and has few
  56 mechanisms to detect failure. Consider the example of such a robot
  57 detecting an obstacle while aiming for a certain goal in its
  58 environment. The system can become deadlocked between competing
  59 behaviours i.e. escaping the obstacle and approaching the goal.
  60 Several simple techniques have been introduced to solve this problem
  61 without resorting to planning, i.e. wandering behaviours [ANDERSON90]
  62 or relocating the goal [ADAMS90]. Nevertheless, the robot has learnt
  63 nothing from its ordeal and will behave in the same inefficient manner
  64 if it encounters the same obstacle again.
  65
  66 This shortcoming has led to the development of hybrid
  67 planning-reactive systems [ARKIN90]. The robot monitors the outcome of
  68 the response to the sensor stimulus and decides if another response
  69 needs to be coupled to this sensor input. In other words, to operate
  70 efficiently a behaviour-based robot must be able to learn to
  71 co-ordinate its behaviours.
  72
  73 The aspect of machine learning is a very interesting and up to date
  74 topic where research is in its infancy. A very promising and popular
  75 approach is the use of so called mobile agents [FRANKLIN96]. The term
  76 "agent" reflects a change of view towards programs; an "agent" has to
  77 be able to cope with broad user expectations. This new view also has
  78 new requirements for machine intelligence, i.e. a user can give the
  79 robot the same orders as he would give to a human. An agent is a
  80 program that acts and decides autonomously in the execution of an
  81 order. In contrast to this, a traditional program is considered to be
  82 passive, as it simply reacts on a user input with a certain output. To
  83 give an example, imagine a hospital with an autonomous mobile robot
  84 that can bring medicine to patients. The doctor tells the robot to
  85 bring a certain medicine to patient X in room Y. The AMR knows where
  86 it can get the medicine and drives to the distribution point. There it
  87 contacts a computer to automatically hand out the medicine. Bad luck,
  88 this specific medicine is not in stock. So the AMR contacts a medicine
  89 database to look for an equivalent product with the same effect which
  90 it can take to the patient. Now just in front of room Y it discovers
  91 that patient X is no longer lying here. A search through the
  92 hospital's database reveals that patient X moved to room Z where the
  93 AMR redirects itself to.  It can easily be seen that classic
  94 approaches would not have been able to solve such a complex task. The
  95 designer of the mobile agent has of course trained the robot to
  96 perform many different behaviours, but the agent makes the decisions
  97 on its own and progressively learns new optimal behaviours.
  98
  99 This concept of mobile agents can be extended to a group of
 100 agents. Members of this group can exchange information to cooperate or
 101 coordinate the next steps in the same task. This requires methods of
 102 communication and some kind of social behaviour, which brings us to a
 103 form of artificial life. A classic example of a task for such a group
 104 of robots is the exploration of an unknown environment.  Each member
 105 of the group explores a section on its own and upon return to a
 106 meeting point merges its information with the information of the other
 107 members to build up a complete map. Another example is the
 108 transportation of an unwieldy object. The robots have to co-ordinate
 109 their actions to find suitable positions around the object to move it
 110 together.
 111
 112 There are widespread applications for an intelligent group of robots
 113 which have not yet been researched. Most of the ideas and techniques
 114 derive from biological models and are implemented by using the
 115 concepts of Artificial Intelligence.
 116
 117
 118 To ease implementation and testing of control programs for mobile
 119 agents, we intend to build a modular and distributed robot simulator.
 120 Agents interacting in a real-world environment have some similarities
 121 to asyncron distributed systems, and it seems obvious that a
 122 distributed system is the best way to simulate this aspect of the
 123 problem domain.
 124
 125 The simulator will be modular to handle new sensors without
 126 recompiling.  The sensor modules will be loaded based on a robot
 127 description file which list the available sensors and the placement on
 128 the robot.  We intent do build modules distance sensors, radio
 129 communication, compass sensor and simulated camera.
 130
 131 The camera simulation will use freely awailable 3D tools to generate
 132 the robot view in real time.
 133
 134 The simulator will be open source and free to use for any other
 135 research project, and we intend to use only freely available libraries
 136 to make sure this system will work on all platforms.
 137
 138
 139 In 1992, Oliver Faugeras published a paper describing a method for
 140 automatic selv-calibration of moving cameras or stereo vision.  We
 141 intend to implement this method on low resolution cameras and low CPU
 142 power equipment.  If we manage to solve the following equation with 9
 143 unknown coefficients in real time, our solution should give very high
 144 frame rate on high CPU computers.
 145
 146  ... XXX fill in equation
 147
 148
 149
 150 Significance
 151 xxxx
 152
 153 Expected outcomes
 154 xxxx
 155
 156 The outcome of the project will be as follows:
 157 - xx,
 158 - xx,
 159 - Simulation system.
 160
 161
 162 17.2 Research plan, methods, techniques
 163
 164 Research plan
 165 xx.
 166
 167
 168 Methods and techniques
 169 xx
 170
 171
 172 17.3 Proposed timing
 173 xx (diagram).
 174
 175
 176 January 2001 to March 2001      xx
 177 April 2001 to June 2001 xx
 178
 179 Robot Operating System
 180 April 2001 to October 2001      xx
 181 November 2001 to March 2002     xx
 182 April 2002 to March 2003        xx
 183
 184 Robot Simulator
 185 April 2001 to September 2001    xx
 186 October 2001 to March 2002      xx
 187 April 2002 to March 2003        xx
 188
 189 Behaviour-Based System
 190 January 2002 to December 2002   xx
 191 January 2003 to July 2003       xx
 192
 193 Evaluation and Improvement
 194 June 2003 to December 2003      Evaluation and Testing
 195 October 2003 to December 2003   Writing of final report
 196
 197
 198 17.4 Justification of the budget
 199
 200 xx.
 201
 202 We need three Ph.D. students to work on this project because ...
 203
 204
 205 17.5 Relevance of applicant skills, training and experience to the project
 206
 207 At Stuttgart Univ., A/Prof. Thomas Bräunl cofounded the "Robotics Lab" with
 208 Prof. Paul Levi. This lab conprised three industrial size mobile platforms, one
 209 carrying an industrial robot manipulator arm, one carrying an innovative stereo
 210 head system. These robot shave been used in several research projects,
 211 including:
 212
 213 - Implementing a robot operating system in the language Oberon
 214 - Tracking and following a leading vehicle
 215 - Object recognition and flexible manipulation
 216 - Cooperation of multiple vehicles in docking and transportation tasks
 217 - Object tracking with the help of a massively parallel stereo camera system
 218 <FIGURE>
 219
 220 At UWA, A/Prof. Thomas Bräunl founded the "Mobile Robot Lab", which comprises
 221 currecntly 15 mobile robot systems of the "EyeBot" family. These are
 222 non-holonomic robots with 2 wheels, holonomic 4-wheel-driven robots, 6-legged
 223 walking robots, and 2-legged walking robots.
 224
 225 <FIGURE>
 226 17.6 Role of each participant
 227
 228 The chief investigator xxx.
 229 The three Ph.D. students will ...
 230
 231
 232 References
 233
 234 [ADAMS90]       MD Adams, H HU and PJ Probert, "Towards a Real-Time Architecture for
 235 Obstacle Avoidance and Path Planning in Mobile Robots", Proceedings of IEEE on
 236 Robotics and Automation Conference, pp584-589, 1990.
 237
 238 [ARKIN90]       RC Arkin, "Integrating Behavioural, Perceptual and World Knowledge in
 239 Reactive Navigation", Robotics and Autonomous Systems, vol.6 pp105-122, 1990.
 240
 241 [ANDERSON90] TL Anderson and M Donath, "Animal Behaviour as a Paradigm for
 242 Developing Robot Autonomy", Robotics and Autonomous Systems, vol.6 pp146-168
 243 1990.
 244
 245 [BROOKS 86]     RA Brooks, "A layered Intelligent Control System for a Mobile
 246 Robot", IEEE Trans. On Robotics and Automation, RA-2, April 1986.
 247
 248 [FRANKLIN96] S Franklin and A Graesser, "Is it an Agent, or just a Program: A
 249 Taxonomy for Autononmous Agents.", Proc.3 Int. Workshop on Agent Theories,
 250 Architecture, and Languages (ATAL), pp 193-206, Budapest 1996
 251
 252 [LATOMBE 91] J-C Latombe, "Robot Motion Planning", Kluwer Academic Press, 1991.
 253
 254 <<more>>