Ausarbeitung/introduction.tex

   1 \chapter{Introduction}
   2 The world nowadays depends more and more on robots. But not only stationary
   3 robots, like those used in assembly lines, improve our lives and make it
   4 possible to excel at industrial and exploratory challenges. Also, mobile robots
   5 could be an integral part of our world in the near future. Not only are they
   6 predestined for dangerous duties which would much certainly cost lives if
   7 humans had to do them, like detecting disarming mines in war zones, they are
   8 also much more specialized to their challenges.
   9
  10 However, in most cases, mobile robots become useless if they don't know anything
  11 about their environment. Especially their location is a basic to know,
  12 and particularly autonomous mobile robots benefit from the knowledge about their
  13 location. This knowledge gives them the ability to create a map of their
  14 environment, to find ways between certain points of interest, and to contact
  15 both human beings as well as other robots.
  16
  17 There are several methods to achieve knowledge about the location of mobile
  18 robots. At first, they can be classified as absolute or relative methods. For
  19 example, absolute positioning methods rely on a fixed, distributed
  20 infrastructure which allows the robot to determine its absolute location inside
  21 a specific region. For example, the \ac{GPS} only works because satellites orbit
  22 around the earth and constantly send a signal. There are also approaches using
  23 navigation beacons or by matching the observed environment to a previously
  24 known map.
  25
  26 All of that positioning methods however depend on the infrastructure, which can
  27 be expensive to deploy and maintain. Also, most of the algorithms used are
  28 fairly complex and need highly specialized hardware or a certain computing
  29 power, and therefore cannot (yet) be used in real time. And sometimes, they are
  30 even more imprecise than relative approaches.~\cite{umbmark}
  31
  32 Relative positioning methods however only yield a position relative to a
  33 defined point, and mostly can be applied with cheap hardware. The best-known
  34 approach in this field is \definition{dead reckoning}, which computes the
  35 position relative to an initial starting point by considering the covered
  36 distance and course the robot has moved. In the case of mobile robots, the
  37 covered distance can be simply computed in real time from the revolution of its
  38 wheels, or by accelerometers the robot may be equipped with.
  39
  40 The advantage of dead reckoning is that it is easy and fast to implement.
  41 However, there is also a drawback. The vehicle's current position is based on
  42 its previous position, and distance and course measurement is mostly imprecise.
  43 the error of the calculated position grows with time. In order to maintain a
  44 certain accuracy, countermeasures need to be taken to keep the error as small as
  45 possible.
  46
  47 In this thesis, the popular Roomba manufactured by the iRobot Corporation is
  48 used as an instance for an mobile robot, since it is cheap, small, and can be
  49 simply controlled over a serial interface. The Roomba is accompanied with a
  50 netbook sitting on top of it, on which runs an application to represent the
  51 autonomous aspect. This application in turn uses the Wiselib library to control
  52 the Roomba.
  53
  54 In the first experiment, the Roomba's movement behaviour will be analyzed. Then,
  55 two different easy approaches for minimizing the positioning error are presented
  56 and tested for their suitability.
  57
  58 \chapter{Related Work}
  59 There are already some results on the problem. Yamauchi proposes to use dead
  60 reckoning with evidence grids to represent the robot's view of the
  61 world.~\cite{yamauchi} Evidence grids are subdivisions of the Cartesian space
  62 into cells, and each cell is assigned a probability value representing the
  63 estimated probability of occupancy by the robot. The robot now explores its
  64 environment, using its sensors to detect moving and fixed obstacles, and finally
  65 tries to find its home location again using dead reckoning. On its way back, it
  66 creates a second evidence grid. Finally, both grids are incorporated by an
  67 algorithm, which estimates the error in orientation and translation.
  68 Experimental results with a real robot showed a maximum translation error of 28\%
  69 and an orientation error of 25\%.
  70
  71 Chong and Kleeman use high-precision odometry to increase the accuracy
  72 of position estimation by dead reckoning.~\cite{chongkleeman} They succeeded in
  73 building a odometry system that is as good as the best reported system, but can
  74 be fabricated at low-cost.
  75
  76 Borenstein and Feng design a benchmark for mobile robots.~\cite{umbmark} They
  77 show how to measure systematic and non-systematic errors, and how to correct
  78 them.