1999
DESIGN, CONTROL, AND APPLICATIONS OF AUTONOMOUS MOBILE ROBOTS
D. FLOREANO, J. GODJEVAC, A. MARTINOLI, F. MONDADA AND J-D. NICOUDMicro-computing Laboratory, Swiss Federal Institute of Technology in Lausanne, LAMI-INF-EPFL, CH-1015 Ecublens
The paper described the Khepera miniature robots developed to support research and developments in research upto that period.
Khepera miniature mobile robot was initially designed in 1991 by E. Franzi, A. Guignard, and F. Mondada, based on ideas of J-D. Nicoud.
F. Mondada, E. Franzi, and P. Ienne. Mobile robot miniaturization: A tool for investigation in control algorithms. In T. Yoshikawa and F. Miyazaki, editors, Proceedings of the Third International Symposium on Experimental Robotics, pages 501{513, Tokyo, 1993. Springer Verlag. 31. A. Murciano and J. del R. Mill an. Lear
Some important features of the robot.
MINIATURISATION
Miniaturisation of a mobile robot brings some advantages to the researcher
HARDWARE AND SOFTWARE MODULARITY
Hardware modularity enables different possible con gurations and experiments using the same basic components. It means also possible extensions and, globally, cheaper equipment. Software modularity means exibility and possibilities for extensions, which enables the software developer to write only parts of the program required for the specifi c application. Khepera is based on this concept of modularity, both in hardware and software.
FROM SIMULATIONS TO APPLICATIONS USING KHEPERA
Several researchers in the field of autonomous mobile robots belong to computer science, arti ficial intelligence. Being a physical robot, it introduces most of the characteristics of robots used for real-world applications. Several Khepera users who have moved from simulations to the miniature robot and have highlighted the advantages of using these modes. Recent construction of the new larger Koala robot, which is software compatible with Khepera, enables the transfer of developments made on the Khepera to a more complex platform which can be used for real-world applications.
Basic con guration
In its basic con guration ( gs. 1 and 2), Khepera consists of two layers corresponding to two main boards: the sensory-motor board and the CPU board. The motor system consists of two lateral wheels and two pivots on the front and back. This con figuration is very good for facing complex geometric obstacles because the robot can turn in place without lateral displacement. The sensory system available in the basic con guration is placed on the lower board, consisting of 8 infrared-light proximity sensors distributed around the body, 6 on one side and two on the other. These
sensors can detect the presence of objects by emitting and measuring reected light and can also be used as simple passive infrared light sensors.
On the sensorimotor board are also placed NiCd batteries with a capacity of 110 mAh which allow the robot to be self-su cient for approximately 30-40 minutes. The CPU board encloses the robot's main processor (a Motorola MC68331 with 128 K-bytes of EEPROM and 256 K-bytes of static
RAM). An A/D converter allows the acquisition of analog signals coming from the sensory-motor board. An RS232 serial line is also available on the board via a miniature connector. On this same connection, a wire can also provide continuous power supply from an external source.
Fuzzy control
Fuzzy logic offers the possibility to express and implement human know-how in the form of linguistic if-then rules which can be applied for the control of nonlinear systems, such as mobile robots. Every rule has two parts: the antecedent part (premise), expressed by If. . . , and the consequent part, expressed by: then. . . . The general form of a linguistic if-then rule is: If a set of conditions is satis ed then a set of consequences can be inferred.
A fuzzy controller is composed of four principal modules. The fuzzi cation interface performs the transformation of crisp values into fuzzy sets. The know ledge base supplies the fuzzi cation module,
the inference engine, and the defuzzi cation interface with necessary information (parameters of membership functions and rules) for their proper functioning. The decision making unit, or inference engine, computes the meaning of the set of linguistic rules. The defuzzi cation interface trans forms the union of fuzzy sets (individual contributions of each rule in the rule base) into a crisp output.
Although one can implement a simple controller for obstacle avoidance on the Khepera with few rules, the main effort is that of designing the appropriate membership functions and choosing the rules.
NEURO-FUZZY CONTROL FOR OBSTACLE AVOIDANCE
Evolutionary Robotics
Evolutionary Robotics is a technique for automatic creation of control systems for autonomous robots that is inspired upon the Darwinian principle of selective reproduction of the fittest individuals.
CO-EVOLUTIONARY ROBOTICS
FULLY AUTONOMOUS ROBOTS APPLICATIONS
A fully autonomous, widely marketed robot is the Husqvarna lawn-mower for at and prepared terrains. The robot is 15 cm high and has a surface of 80 cm by 40 cm covered by solar power cells which let it work for several months when the sun is high over the sky (a small battery back-up is used by the processor when sun light is not strong enough to power the robot). The robot moves randomly, exploiting small irregularities of the terrain, while checking for an electric wire (solar powered too) positioned on the perimeter by the owner. In case the robot gets stuck in unexpected situations, it starts beeping and waiting for human help. Lawn-mowing is a simple navigation
task where random walk seems acceptable; furthermore, since the wheels move faster when the lawn is cut, the robot tends to spend more time on areas not yet cleared. Pool cleaning robots share some characteristics with the autonomous lawn-mower. Although several types are available on
the market (e.g., see http://h2o-marketing.com/aquabot/aqua.html),
they generally perform a random walk on the bottom and on the sides
while scrubbing, vacuum cleaning, and ltering the pool. They are generally
powered via a cable hanging from the centre of the pool and can also
be remotely controlled, if necessary. A more systematic cleaning of the
pool inner surface can be achieved by pressure sensors which exploit the
regularities of tiles.
Currently, autonomous vacuum cleaning robots are restricted to speci FIc large environments, such as airport lounges (Narita airport in Japan, e.g.). A prototype robot in the Paris metro was designed to follow a line buried in the ground, whereas the recently completed CLEAN Eureka pro ject (nr. EU-1094 on the Eureka database: http://www.eureka.be) has attempted to develop a robot for
cleaning hyper-market surfaces by exploiting pre-positioned active landmarks.
Recently, a
legged water-proof robot has been developed for landmines positioned on the surf zone, which is a rather regular and de ned terrain. All these robots are supposed to blow up mines by hitting them (and being destroyed in the meanwhile).
The Mars So journer (http://mpfwww.jpl.nasa.gov/default.html) is a popular example of a robot with full energy autonomy, but limited behavioural autonomy. Since it moves very slowly, a rather small area of solar cells is sufficient to power the robot. It receives instructions from Earth on
its destination, but it has to get there autonomously.
Another application with similar behavioral requirements is the autonomous wheelchair. Several handicapped persons find it difficult to steer precisely their own wheelchair to get around corners or passing through doorways. By supplying these chairs with additional sensors and appropriate control systems that support semi-autonomous navigation, the owner can instruct the chair on the desired destination and let it get there autonomously.
Semi-autonomous mobile robots have a large potential market, from rescue robots to robots for maintenance of nuclear plants, and have several military applications, such as reconnaissance
ying drones.
An application that has attracted the interest of several industries, research institutes, and funding agencies is a semi-autonomous vehicle capable of navigating in daily tra c as well as on rough terrains. A well-known example is NavLab [20], developed by Carnegie Mellon University. The Swiss Serpentine is a urban semi-autonomous vehicle designed for accommodating several
standing persons which follows an inductive track providing power, self-localisation, and general directives on the task to be accomplished.
No comments:
Post a Comment