1 00:00:00,780 --> 00:00:07,470 Previously we tested our measles on the first symptoms and it passed the test with flying colours. 2 00:00:07,890 --> 00:00:15,210 Can the missile will do the same if we provided a slightly tougher opponent, a complex maze with sharp 3 00:00:15,210 --> 00:00:19,080 tones and irregular walls, will it be able to solve it? 4 00:00:19,830 --> 00:00:20,770 Let's find out. 5 00:00:21,450 --> 00:00:24,360 To test our missile robot on the more complex maze. 6 00:00:24,360 --> 00:00:27,830 Two, we already have the maze to burn inside the world. 7 00:00:27,880 --> 00:00:28,380 Follow it. 8 00:00:28,740 --> 00:00:30,480 And here we have the means to work. 9 00:00:30,840 --> 00:00:32,520 So all we need to do to test it out. 10 00:00:32,790 --> 00:00:33,930 Our order at home will. 11 00:00:34,870 --> 00:00:41,440 Build the project and then launch the space to simulation, and that is north by simply accessing the 12 00:00:41,440 --> 00:00:48,130 maze to robot camera dot launched by instead of the maze one robot camera spotlight or by this starts 13 00:00:48,130 --> 00:00:52,000 up the maze two or the more complex maze two simulation. 14 00:00:52,630 --> 00:00:59,050 And once it starts up, you will notice that this particular maze is a lot more complex than the maze 15 00:00:59,050 --> 00:00:59,290 one. 16 00:00:59,710 --> 00:01:06,760 It has sharp buttons and irregular bolts, so a lot more tougher problem for our solver board to solve. 17 00:01:07,270 --> 00:01:11,500 So let's head on over to the second terminal and run the maze solver. 18 00:01:13,250 --> 00:01:17,990 So what we get as the first output is basically the output of the localizing stage where we have the 19 00:01:17,990 --> 00:01:20,330 occupancy grid computed for our maze. 20 00:01:20,330 --> 00:01:26,310 Two And as you can see, this seems to be the correct occupancy grid showing us the areas the car can 21 00:01:26,330 --> 00:01:26,640 go. 22 00:01:26,750 --> 00:01:29,780 In this particular case, pressing the spacebar. 23 00:01:31,890 --> 00:01:33,240 We get the falling out. 24 00:01:33,570 --> 00:01:37,230 That is the life may solving done by our missile robot. 25 00:01:37,740 --> 00:01:43,590 So as you can see, the car has moved from point A to point B, where it has computed is correct or 26 00:01:43,590 --> 00:01:44,400 complete robot. 27 00:01:44,910 --> 00:01:48,360 And then it goes back and tries to lose the first mini. 28 00:01:48,360 --> 00:01:51,090 Go for our short a spot to go. 29 00:01:51,570 --> 00:01:59,400 So let's observe the right frame of reference where we have drawn the shortest path to go and also the 30 00:01:59,400 --> 00:02:02,850 current mini goal and the mini goal that we have already completed. 31 00:02:03,240 --> 00:02:07,950 So you can see the orange, mini orange basically circles shows the mini goal that we are trying to 32 00:02:07,950 --> 00:02:08,310 achieve. 33 00:02:08,490 --> 00:02:11,490 And the green mini goal is basically what we have already done. 34 00:02:12,150 --> 00:02:13,620 So, so far, so good. 35 00:02:13,830 --> 00:02:20,880 Our missile abort is able to solve this particular maze correctly and is trying to achieve or reach 36 00:02:20,880 --> 00:02:23,130 this particular meaning goal, which is very far from it. 37 00:02:23,310 --> 00:02:29,040 So it moves very fast and then it approaches its first major obstacle. 38 00:02:29,820 --> 00:02:33,990 Our missile abort is approaching a very narrow corridor. 39 00:02:34,380 --> 00:02:40,470 So this is the corridor where our missile robot might get stuck because of its limited algorithm. 40 00:02:41,160 --> 00:02:49,800 So it is approaching the mini goal one after another, and then it tries to reach the corridor and then. 41 00:02:51,040 --> 00:02:54,160 We get the many go that is somewhere on the wall. 42 00:02:54,850 --> 00:02:57,400 So what is a molecule on the wall? 43 00:02:57,670 --> 00:03:00,310 Because the part should never be on the wall. 44 00:03:00,700 --> 00:03:06,100 The problem is that with the image that we are getting is prospectively distorted because there is a 45 00:03:06,100 --> 00:03:09,400 problem that can occur with cameras and the images. 46 00:03:09,910 --> 00:03:15,640 So with this problem, we are getting that part computed somewhere on the wall and which is the area 47 00:03:15,700 --> 00:03:17,800 which our missile robot cannot reach. 48 00:03:18,160 --> 00:03:23,860 So for such a case, the car would try to reach over that wall, which it cannot, and get stuck in 49 00:03:23,860 --> 00:03:24,940 that particular position. 50 00:03:25,270 --> 00:03:30,850 So if you observe the car in the simulation, you will see that it is stuck behind the wall and trying 51 00:03:30,850 --> 00:03:31,960 to move towards the wall. 52 00:03:32,530 --> 00:03:38,890 The reason is that the computer you go is somewhere on the wall and it is incorrectly computed and that 53 00:03:38,890 --> 00:03:41,080 is why the car is stuck right at this moment. 54 00:03:41,650 --> 00:03:47,170 So we need to tell our to go that it is if it is stuck behind the wall and the mini goal is somewhere 55 00:03:47,170 --> 00:03:51,190 on the obstacle, try to reach the next minute goal that is not on the wall. 56 00:03:51,640 --> 00:03:57,550 And the same way, if the car gets stuck on some other wall that is irregular or is a relationship that 57 00:03:57,550 --> 00:04:00,280 it needs to try the next appropriate goal. 58 00:04:01,060 --> 00:04:06,850 So we need to create some sort of obstacle navigation so that the car can reach the exit. 59 00:04:06,970 --> 00:04:08,660 In this more complex means too. 60 00:04:09,210 --> 00:04:10,210 Let's try and do that.