Localization & Location-Based Services with Particle Filters 2010 - Revision history

Robot Lab 2010: /* Software */

2010-03-19T23:57:27Z

Software

Robot Lab 2010: /* Software */

2010-03-19T23:57:07Z

Software

Robot Lab 2010: /* Software */

2010-03-19T23:55:24Z

Software

Robot Lab 2010: /* WiFi Localization Technique */

2010-03-19T23:55:08Z

WiFi Localization Technique

Robot Lab 2010: /* Algorithms */

2010-03-19T23:54:59Z

Algorithms

Robot Lab 2010 at 23:54, 19 March 2010

2010-03-19T23:54:43Z

Robot Lab 2010: /* Emerging solutions */

2010-03-19T23:53:57Z

Emerging solutions

Robot Lab 2010 at 23:53, 19 March 2010

2010-03-19T23:53:29Z

Robot Lab 2010 at 23:50, 19 March 2010

2010-03-19T23:50:55Z

Robot Lab 2010 at 23:50, 19 March 2010

2010-03-19T23:50:10Z

@@ Line 271: / Line 271: @@
 I made minor changes to the aima core library. One example of a minor change was to stop looping infinitely, when there were not enough particles found.
 [http://www.cs.drexel.edu/~maj39/cs511/aima-core-test-changed.jar http://www.cs.drexel.edu/~maj39/cs511/aima-core-test-changed.jar]
-[[http://www.cs.drexel.edu/~maj39/cs511/aima-core-changed.jar http://www.cs.drexel.edu/~maj39/cs511/aima-core-changed.jar]]
 The NetStumbler wifi scan files are available here: [http://www.cs.drexel.edu/~maj39/cs511/Scans.zip http://www.cs.drexel.edu/~maj39/cs511/Scans.zip]

@@ Line 271: / Line 271: @@
 I made minor changes to the aima core library. One example of a minor change was to stop looping infinitely, when there were not enough particles found.
-<aima-core-test-changed.jar> <aima-core-changed.jar>
+[http://www.cs.drexel.edu/~maj39/cs511/aima-core-test-changed.jar http://www.cs.drexel.edu/~maj39/cs511/aima-core-test-changed.jar]
 The NetStumbler wifi scan files are available here: [http://www.cs.drexel.edu/~maj39/cs511/Scans.zip http://www.cs.drexel.edu/~maj39/cs511/Scans.zip]

@@ Line 263: / Line 263: @@
 This is the video of the Particle filter in action. [http://www.cs.drexel.edu/~maj39/cs511/OutSide.swf http://www.cs.drexel.edu/~maj39/cs511/OutSide.swf]
 == '''Software''' ==
@@ Line 273: / Line 274: @@
 The NetStumbler wifi scan files are available here: [http://www.cs.drexel.edu/~maj39/cs511/Scans.zip http://www.cs.drexel.edu/~maj39/cs511/Scans.zip]
 == '''Implementation Details''' ==

@@ Line 167: / Line 167: @@
 * Completely passive localization is possible by "sniffing" wireless traffic
 * WiFi signals do not require line-of-sight. As such, they can be used even in the presence of obstructions that would block laser range-finders or cameras.
 == '''WiFi Localization Technique''' ==

@@ Line 151: / Line 151: @@
 There are two problems that need to be solved: Localization and Path finding.  This project is going to attempt the Localization part only.
 '''Localization'''
@@ Line 166: / Line 167: @@
 * Completely passive localization is possible by "sniffing" wireless traffic
 * WiFi signals do not require line-of-sight. As such, they can be used even in the presence of obstructions that would block laser range-finders or cameras.
 == '''WiFi Localization Technique''' ==

@@ Line 123: / Line 123: @@
 * The time taken by the user to reach its destination is a factor of the time taken by the robot to navigate around obstacles.
 * creates possibility for extra work, i.e to make the robot socially acceptable, as compared to using already acceptable devices such as hand held smart-phones.
 == '''My Solution''' ==
@@ Line 143: / Line 144: @@
 This application has the following disadvantage:
 Visitors need to have a smartphone / mobile computing device in order to use this solution.
 == '''Algorithms''' ==
@@ Line 174: / Line 176: @@
 The signal strength is a function of distance and the obstacles that might be in between the robot and the origin of these signals i.e. the Access Points (APs)  One obvious solution could be triangulation of the signal strengths from multiple Access Points (APs).  But unfortunately the signal strength is a complex function of distance, because there is a lot of noise on the wireless channel in indoor environments and the radio frequency (RF) signal can suffer from reflection, diffraction and multipath effect.  To overcome this issue, a radio map can be created, which captures the signatures of each Access Point (AP) at certain points in the area of interest.
@@ Line 389: / Line 390: @@
 Although looking at the chart for the Drexel's Computer Science Department's Corridor, it can be deduced that the Particle Filter algorithm did not give very good results.  But upon viewing the attached video sequence, one can see that the particles are not far behind, or are in the corridor's vicinity.  The problem here is the environment, since the thin long Corridors can propagate the Signals easily, that can cause the signal's strength to stay unchanged or change by a very little amount.
 == '''Future Work''' ==
@@ Line 398: / Line 400: @@
 '''Physical walls, barriers model'''
 We can take into consideration the fact that people can not walk through walls or other physical barriers.  We can use this fact to refine the results given by the particle filters.

@@ Line 46: / Line 46: @@
 Following lists some of the new ideas that researchers have been working on to solve this problem.
 '''ClickAndGo Way Finding Maps from Intouch Graphics'''
@@ Line 62: / Line 63: @@
 * Service is for public facilities only, it does not cover private buildings such as office buildings or conference halls.
 * Data is prepared by specialists.  Comparatively with automated techniques, this can be more time consuming.
 '''LAIR (Location Awareness Information Representation)'''
@@ Line 74: / Line 76: @@
 * Not able to provide with information relevant to user's current location.
 '''Handheld Campus Tour Guide'''
@@ Line 88: / Line 91: @@
 * Although this project describes work which is very similar to my intended idea, it is not clear if the project met its intended goal.
 '''SafeLink'''
@@ Line 103: / Line 107: @@
 * It is a commercial service.
 * Uses proprietary devices.
 ''''LUCAS', Limerick University Computerized Assistive System'''
@@ Line 118: / Line 123: @@
 * The time taken by the user to reach its destination is a factor of the time taken by the robot to navigate around obstacles.
 * creates possibility for extra work, i.e to make the robot socially acceptable, as compared to using already acceptable devices such as hand held smart-phones.
 == '''My Solution''' ==

← Older revision		Revision as of 23:53, 19 March 2010
Line 1:		Line 1:
−	~~By Ahsen Jaffer~~
−
−
−

	== '''Description''' ==		== '''Description''' ==

← Older revision		Revision as of 23:50, 19 March 2010
Line 214:		Line 214:


−	This is the video of the Particle filter in action. [~~[Media~~:Library.~~mov]~~]	+	This is the video of the Particle filter in action. [http://www.cs.drexel.edu/~maj39/cs511/Library.swf http://www.cs.drexel.edu/~maj39/cs511/Library.swf]

← Older revision		Revision as of 23:50, 19 March 2010
Line 128:		Line 128:

	An automated assistant in the form of an application that can be installed on to the visitors' smartphone / mobile computing device. The visitors will be able to use this automated assistant application while walking around the building. This automated assistant application will know its current location and also be able to give instructions to navigate them through the building.		An automated assistant in the form of an application that can be installed on to the visitors' smartphone / mobile computing device. The visitors will be able to use this automated assistant application while walking around the building. This automated assistant application will know its current location and also be able to give instructions to navigate them through the building.
		+

	''' Why is this a good idea?'''		''' Why is this a good idea?'''