These are corrections to the book
Computational
Principles of Mobile Robotics
by Dudek and Jenkin.
Further corrections or suggestions are welcome. They can be
mailed to dudek@cim.mcgill.ca
- p. 31, Figure 2.14: the robot pictures is called the
Honda P2. It is
not rho-2. A more recent version of the robot, called P3, also
exists and looks roughly the same the the one shown.
In more recent developments, Sony also has a (smaller) humanoid
robot.
-
p.65, Fig.3.26: The caption should read (c) Spray, and (d) Theseus. (Thank to Artur Wolek for spotting this.)
- p. 78, Figure 3.12 (EKF example), line 4:
the derivation of Lambda-E is erroneous. It is based on the
partial derivative matrix of h, but based on the expression for
h used on p.74 there is a denominator term missing
in each row of the
matrix. This should be like the original expression for h but to the
power (-1/2) instead of (1/2). (In other words, we accidentally dropped
a term when transcribing the derivative.)
- p. 230, fifth line in 8.4.1: "Rendzevous" should
be "Rendezvous".
- p. 212, sixth line should read "alterations", not
"alternations".
- p. 217, last paragraph reads:
"...pose is given in a 2-D environment by q = <x,y,theta>,
where x and y specify orientation and theta
specifies orientation of the robot."
it should read
"...pose is given in a 2-D environment by q = <x,y,theta>,
where x and y specify the position
and theta specifies the orientation of the robot."
- p. 146: there is a reference to the "Randomized Path Planner (RPP) discussed
in section 5.3.3". A more helpful backwards
pointer might be "the Randomized Path
Planner (RPP) discussed on page 140."
Additionally, this paragraph comments on the shortcomings of early
RPP work for complex confugration spaces.
It should be emphasised that the notion
of a randomized (or probabilistic)
path planner has been substantially elaborated and
can now be used very effectively particularly for configuration spaces
with a very large number of degrees of freedom. Reference 181 is
an excellent example of this kind of work.
- p.56: Hooke's law for the restoring
force exerted by an ideal spring is directly proportional to the amount
it has been compressed, thus
F=kx, and the potential energy stored by the spring is thus E=(kx^2)/2,
thus we have a=kx/m for a spring-like accelerometer. The displacement of the
mass is thus proportional to the acceleration x=ma/k.
- Update: The Glonass system developed by the Soviet Union, and now
inherited by Russia, has not been keeping up with Navstar.
Many of the Glonass system's Uragan satellites have failed and the system
has 12 or 13 satellites in orbit, although the system does
not appear to be functional as of April 2001. There have been claims that the
system can be made partly operational with 12 satellites until 2005, and that it will
eventually be replaced or upgraded by a system referred to as Glonass-M.
An official GLONASS
web page might provide further details.
- p. 181 last paragraph:
"For example, in the context of (Section 3.8), GDOP is sometimes used ..."
the word GPS is missing before the word 'of' before the paranthesis. i.e.
the sentence should read
"For example, in the context of GPS (Section 3.8), GDOP is sometimes used ..."
- In section 3.4 on infrared sensors, the 3rd
paragraph of page 59 begins with "The process of inferring accurate
distance information from an infrared sonar ring is nontrivial ..."
The word "sonar" should be replaced by "sensor".