| 0:00:00 | hi my name is william attacker gonzalez and i'm doing a project on a time |
|---|
| 0:00:04 | to miss navigation and control |
|---|
| 0:00:06 | evaluating methods to simultaneous localisation and mapping or slam |
|---|
| 0:00:12 | as you may have heard and then use recently |
|---|
| 0:00:15 | there was a recent |
|---|
| 0:00:16 | playing flight |
|---|
| 0:00:17 | motion plane flight crash |
|---|
| 0:00:19 | number three seventy |
|---|
| 0:00:20 | where hundreds of people are gone missing |
|---|
| 0:00:23 | and right now sciences are using a time to miss underwater vehicles in a huge |
|---|
| 0:00:29 | search operation |
|---|
| 0:00:30 | using the technology known as slam to map |
|---|
| 0:00:34 | and visualize the motion for to look for displaying plate |
|---|
| 0:00:39 | so what is slam |
|---|
| 0:00:41 | after getting little background so i am it is simultaneous localisation mapping as i said |
|---|
| 0:00:46 | before |
|---|
| 0:00:47 | and it uses complex geometric form as in sure gonna metric functions for robot to |
|---|
| 0:00:53 | scan visualise environment around it as well as position itself so it can know where |
|---|
| 0:00:58 | it is relative to all the objects surrounding it |
|---|
| 0:01:03 | so some approaches to slam better being used today |
|---|
| 0:01:06 | our sensor integration where robot will have not only just a camera to scan |
|---|
| 0:01:12 | by the infrared sensor all just sighing sensor et cetera et cetera |
|---|
| 0:01:16 | and uses hierarchical methods |
|---|
| 0:01:19 | to help the robot's can |
|---|
| 0:01:21 | environment using probability theory |
|---|
| 0:01:24 | so let's say the robot is gaining a glass door obviously the camera is not |
|---|
| 0:01:29 | going to return |
|---|
| 0:01:31 | the best |
|---|
| 0:01:32 | quality data so the way the ultrasonic sensor over the camera |
|---|
| 0:01:38 | so getting in a my product might pieces was as a tandem since robot such |
|---|
| 0:01:44 | as drivers cars are use a novel environments |
|---|
| 0:01:47 | scanning images of an environment using fast scanning algorithms |
|---|
| 0:01:51 | we improve upon the efficiency of mapping as compared with the current standard methods |
|---|
| 0:01:57 | and an overview of my project |
|---|
| 0:02:00 | i did a simulation of slam algorithms and scanning speed |
|---|
| 0:02:04 | using different sensors |
|---|
| 0:02:06 | that were working together |
|---|
| 0:02:08 | and i would evaluate these maps using mahalanobis distance equation |
|---|
| 0:02:13 | which is as shown below |
|---|
| 0:02:16 | so the initial goals and objectives of my product work to take a mobile robot |
|---|
| 0:02:20 | such as nine robot remember |
|---|
| 0:02:22 | and they connect the x box sensor |
|---|
| 0:02:25 | and i would but there was together |
|---|
| 0:02:28 | couple them together you know these are two things you can get |
|---|
| 0:02:31 | really well priced at best buy |
|---|
| 0:02:33 | local best buy |
|---|
| 0:02:35 | to navigate environment and map it |
|---|
| 0:02:37 | and i whatever analyze these physical environment s |
|---|
| 0:02:42 | so my experimental process involved using online open source code from |
|---|
| 0:02:48 | something on is "'cause" e bow |
|---|
| 0:02:50 | i would then start of virtual world |
|---|
| 0:02:53 | and one minute i missed mode in that virtual world |
|---|
| 0:02:56 | and start mapping and visualising |
|---|
| 0:02:59 | now after that was done i would analyse my results |
|---|
| 0:03:03 | so here's my project set up this is all on my laptop i did a |
|---|
| 0:03:06 | few test with the physical environment but unfortunately i wasn't able to show those here |
|---|
| 0:03:11 | but issue here on the left side you can see an odometer a compiler |
|---|
| 0:03:15 | where the robot |
|---|
| 0:03:17 | is very ago ready to start scanning |
|---|
| 0:03:19 | it's ready to localise itself |
|---|
| 0:03:21 | from its initial position by keeping track of |
|---|
| 0:03:24 | how far as wheels have spun |
|---|
| 0:03:27 | and on the right is the complete scans of a simple q the bottom when |
|---|
| 0:03:31 | action a cylinder and top right |
|---|
| 0:03:34 | as you can see is the speed increases |
|---|
| 0:03:38 | for a relatively |
|---|
| 0:03:40 | the |
|---|
| 0:03:42 | quality of the maps goes down |
|---|
| 0:03:44 | as it approaches the right picture |
|---|
| 0:03:47 | and my results here's a graph |
|---|
| 0:03:49 | on the bottom right bar graph |
|---|
| 0:03:51 | showing them how nervous |
|---|
| 0:03:53 | value |
|---|
| 0:03:55 | verses the speed of the scanning |
|---|
| 0:03:58 | mahalanobis value when it's a value of zero means is a perfect scan so my |
|---|
| 0:04:02 | very low speeds can almost return a value of zeros about twenty one i was |
|---|
| 0:04:07 | terry surprise how well that when i'm |
|---|
| 0:04:09 | as you can see the mahalanobis five gets higher and higher |
|---|
| 0:04:12 | and efficiency goes down with the bowser scanning speed |
|---|
| 0:04:16 | which unfortunately i did |
|---|
| 0:04:20 | did not think would happen |
|---|
| 0:04:23 | so significance of my project it's a very cheap alternative to vary |
|---|
| 0:04:27 | complex technology being used today the google car actually |
|---|
| 0:04:32 | which |
|---|
| 0:04:32 | self navigate to self use the technology notice light are |
|---|
| 0:04:36 | which is hunters the thousand dollars |
|---|
| 0:04:38 | so this was you know open source project that i did |
|---|
| 0:04:41 | it's something simple |
|---|
| 0:04:44 | that high school students like me can do for |
|---|
| 0:04:47 | a very small amounts of money |
|---|
| 0:04:50 | very cheap |
|---|
| 0:04:51 | and you know i was very surprise of the quality of maps |
|---|
| 0:04:53 | that my programs generated |
|---|
| 0:04:55 | from doing this project |
|---|
| 0:04:57 | and that's it thank you |
|---|