[Cleaning]

Author: martinferianc

Diff for: README.md

@@ -1,24 +1,17 @@
 # Robotics-EIE3
-Repo to hold the the code for Robotics course CO333 at Imperial
+Repo to hold the the code for Robotics course CO333 at Imperial College London taught by Andrew Davison: [Link](https://www.doc.ic.ac.uk/~ajd/Robotics/)
 
-## **Note**
-Motor port 3 is fried! Do not use it.
+### Contents
 
-## SSH from DOC computers, IP agnostic method:
-`ssh-pi 44:33:4c:6c:c1:46`
+During several weeks we were encouraged to complete several challenges in hadrwarde and software calibration of our proposed robot.
 
-Password: mmoggmmo
+The challanges included:
 
-Node.js interface password: gmmommog
+1. Building the robot and calibratiing it's odometry and PID controller [asssignment_1](/assignment_1).
+2. Observe possible sensors and program avoidance manouvers based on ultrasound readings and bumpers [asssignment_2](/assignment_2) and [asssignment_3](/assignment_3).
+3. Applying particle filter via Monte Carlo algorithm to navigate the robot through a a pre-defined maze [asssignment_4](/assignment_4)
+4. Trying to learn a particular signature of a location to perform continuous localisation [asssignment_5](/assignment_5).
+5. Final challenge in avoiding obstacles to reach a final destination [challenge](/challenge).
 
-MAC Address: 44:33:4c:6c:c1:46
-
-## Accessing Pi from web
-[https://www.doc.ic.ac.uk/~jrj07/robotics](https://www.doc.ic.ac.uk/~jrj07/robotics)
-
-type in MAC address: 44:33:4c:6c:c1:46
-
-## Running scripts
-Run all scripts from the root directory.
-e.g.
-`python test_scripts/test_wheels.py 20 20`
+### Authors:
+George Punter, Owen Harcombe, Mike Chen, Martin Ferianc at Imperial College London 2017.

Diff for: assignment_5/README.md

@@ -1,2 +1,13 @@
 # Robotics-EIE3
-## Robotics Assignment 4:
+## Robotics Assignment 5: Continuos localisation
+[**The handout**](../Resources/questions6.pdf)
+
+An alternative relocalisation technique to monte carlo localisation involves making a lot of measurements at a number of chosen locations and learning their characteristics.
+
+• This can be done without a prior map but needs training
+• The robot can only recognise the locations it has learned
+
+First the robot must be placed in each target location to learn its
+appearance
+
+The raw measurements are stored to describe the location: a place descriptor, or signature. Hence after each rotation of e.g.: an ultra sound sensor it is possible to take it's measurement and record it, later it is possible to compare it with another measurement to determine the current position. To save computation power it is possible to record just the raw distances in a histogram but a disadvantage it the lack of identification of the curent pose.

Diff for: challenge/README.md

@@ -0,0 +1,10 @@
+# Robotics-EIE3
+## Robotics Challenge: Obstacle avoidance
+[**The handout**](../Resources/questions7.pdf)
+
+The aim of the challenge was to navigate across an area filled with obstacles. The robot will
+start with no knowledge of the positions of the bottles and will have to locate them with its sonar and plan a path to advance as far and as quickly as possible.
+
+The robot started behind a start line marked on the floor, facing forwards, and there was a  parallel goal line 3m ahead.
+
+The obstacles we used are soda bottles, which were considered as heavy vertical cylinders with radius 6cm.

Diff for: src/robot.py

@@ -479,18 +479,14 @@ def set_ultra_pose(self, pose):
     def set_robot_pose(self, s_pose, update_particles = False):
         success = True
         print("Starting pose: {}".format(self.state["pose"].get("theta",-1)))
-
         while s_pose >= 360:
             s_pose-=360
         while s_pose <= -360:
             s_pose+=360
-
         rotation = (s_pose-self.state["pose"].get("theta", 0))
-
         if rotation==0:
             print("No rotation required.")
             return True
-
         if rotation > 180:
             rotation-=360
         elif rotation < -180:
@@ -518,7 +514,6 @@ def interactive_mode(self):
                 print("Enter pose to rotate to:")
                 s_pose = float(input())
                 self.set_robot_pose(s_pose)
-
             elif command == 3:
                 distances = []
                 print("Enter left wheel distance:")