Showing Revision 1 created 06/29/2012 by Amara Bot.

1. Title:
   06-08 Your Robot Car
2. Description:

   Other units in this course below:
   Unit 1: http://www.youtube.com/playlist?list=PL1EF620FCB11312A6
   Unit 2: http://www.youtube.com/playlist?list=PL107FD47786234011
   Unit 3: http://www.youtube.com/playlist?list=PL5493E5D24A081719
   Unit 4: http://www.youtube.com/playlist?list=PLAADAB4F235FE8D65
   Unit 5: http://www.youtube.com/playlist?list=PL1B9983ACF22B1920
   Unit 6: http://www.youtube.com/playlist?list=PLC9ED5AC39694C141
   QA: http://www.youtube.com/playlist?list=PL3475310BFB1CBE34

   To gain access to interactive quizzes, homework, programming assignments and a helpful community, join the class at http://www.udacity.com

3. 0 0:00 - 0:06
   Now let's put them all together into a single piece of software.
4. 6000 0:06 - 0:09
   Upfront it took me about a whole day to do this.
5. 9000 0:09 - 0:12
   I'm not going to ask you to do it all yourself,
6. 12000 0:12 - 0:16
   because it's going to cost you probably at least an hour if it takes me a day.
7. 16000 0:16 - 0:22
   But I still want to be able to take all the lessons that we did together into a single system.
8. 22000 0:22 - 0:25
   I'm going to help you a little bit--bits and pieces--
9. 25000 0:25 - 0:29
   but up front here is the environment that I wrote for you,
10. 29000 0:29 - 0:33
    which is very much derived from the environment we studied in the past.
11. 33000 0:33 - 0:37
    We have a class robot that has certain kinds of noise characteristics you can find over here.
12. 37000 0:37 - 0:47
    As I scroll down, you can see the familiar init function, the position-setting function,
13. 47000 0:47 - 0:54
    the set_noise function, and then we have two checking functions--
14. 54000 0:54 - 0:59
    whether we have a collision with the world called "grid," which I will show you in a minute.
15. 59000 0:59 - 1:02
    and we have a check_goal function to see if we reached a goal
16. 62000 1:02 - 1:05
    according to a certain distance threshold.
17. 65000 1:05 - 1:11
    The move function should be very familiar at this point.
18. 71000 1:11 - 1:15
    It applies noise to the motion command.
19. 75000 1:15 - 1:20
    It the same code that you originally wrote.
20. 80000 1:20 - 1:24
    Then we have a very simple sense function, which measures the robot's (x, y) location,
21. 84000 1:24 - 1:32
    similar to a GPS on a car but with substantial measurement noise.
22. 92000 1:32 - 1:36
    Corresponding to this sense function we have a measurement probability function
23. 96000 1:36 - 1:40
    that you might want to use in your filter, and it evaluates the probability
24. 100000 1:40 - 1:47
    of a measurement relative to the ground truth coordinates of the robot using Gaussians.
25. 107000 1:47 - 1:50
    Armed with all this here is the problem.
26. 110000 1:50 - 1:55
    I'm going to give you a grid. Here is an example grid. Let me draw this for you.
27. 115000 1:55 - 1:59
    This specific one happens to be of dimensions 6 and 5,
28. 119000 1:59 - 2:03
    and there are a number of blocked off cells like this.
29. 123000 2:03 - 2:09
    If we look carefully into the code, you'll find information about the initial starting location
30. 129000 2:09 - 2:14
    on the left upper corner and the goal location, which is the bottom right corner.
31. 134000 2:14 - 2:18
    In putting everything together, you're going to build a robotic car using our bicycle model
32. 138000 2:18 - 2:22
    that'll drive through the free space through the continuous free space
33. 142000 2:22 - 2:28
    on something close to the shortest path all the way into the goal.
34. 148000 2:28 - 2:31
    Here is my solution that I implemented
35. 151000 2:31 - 2:35
    and that you will get to see for the most part towards the end of this class.
36. 155000 2:35 - 2:38
    I am starting up over here. These are my obstacles.
37. 158000 2:38 - 2:43
    They implement as circles through the center of these grid cells.
38. 163000 2:43 - 2:47
    It's not exactly correct but good enough for my implementation.
39. 167000 2:47 - 2:51
    Here are multiple runs using the same code, and you can see they're far from optimal.
40. 171000 2:51 - 2:53
    They are non-optimal because there is control noise,
41. 173000 2:53 - 2:55
    and there is also measurement noise.
42. 175000 2:55 - 3:02
    But they all make it safely through free space into the corner where the goal objective is.
43. 182000 3:02 - 3:07
    If we look at them in detail, like this solution over here.
44. 187000 3:07 - 3:11
    You'll find that the spacing of the circles is somewhat variable.
45. 191000 3:11 - 3:15
    You'll find that there's little corners over here that are either the result of control noise
46. 195000 3:15 - 3:20
    or of measurement noise or of my somewhat deficient implementation.
47. 200000 3:20 - 3:24
    You'll also find the control set points that are the smooth points of my A-star planner
48. 204000 3:24 - 3:26
    as shown here in green.
49. 206000 3:26 - 3:31
    In the version that I implemented for you, the controller does something very, very different.
50. 211000 3:31 - 3:35
    It actually chooses as the control objective to head straight to the goal,
51. 215000 3:35 - 3:40
    using the atan2 function, executes the action at a speed of 0.1,
52. 220000 3:40 - 3:44
    and then reports a collision whenever the robot is moving.
53. 224000 3:44 - 3:48
    Just looking down to the output where we see the robot's coordinates
54. 228000 3:48 - 3:52
    along with the orientation there are very frequent collisions
55. 232000 3:52 - 3:56
    that the robot undergoes in its attempt to reach the goal,
56. 236000 3:56 - 4:01
    which it eventually does, but you can see 2 big regions of collisions
57. 241000 4:01 -
    until the goal is finally reached.