Lab #1 Introduction to AMR’s iRobot Create

MAE 4180/5180, ECE 4772/5772, CS 3758
Autonomous Mobile Robots: Lab #1
Introduction to AMR’s iRobot Create

Objectives
In this lab, students will:
• Control the iRobot Create platform
• Change between coordinate frames
• Model sensor error
• Observe actuation error
• Observe the timing effects of obtaining different sensor information
• Perform feedback linearization for controlling a differential drive robot
This lab is designed to familiarize students with the iRobot Create platform that will be used for the remainder of the course. The Create is a differential drive (non-holonomic) robot equipped with several different sensors. For homework assignments and testing outside the lab, a simulator has been developed so
you can debug your code without a physical robot. The simulator is designed to use the same functions as
the MATLAB Toolbox for the iRobot Create (pages 10–11 of Matlab Toolbox iRobot create doc.pdf
posted on Canvas).
Important note about saving files: Make sure to save your data frequently to a shared google drive,
or better yet, have your working Matlab directory be on the USB drive.
Required Code (from Homework 1) and plots (to be shown to
the TAs at the beginning of the lab)
• turnInPlace.m
• readStoreSensorData.m
• OpenLoopControl.m (From the Lab 1 assignment on Canvas)
• backupBump.m
• limitCmds.m
• visitWaypoints.m
• feedbackLin.m
• A plot of the robot trajectory when following waypoints [-1 0; 1 0] (last part of the question about
Feedback Linearization)
1 Lab Manual
1.1 Set-up - Remote student
(a) Join the lab Zoom session (link on Canvas).
(b) Open the Twitch stream: https://www.twitch.tv/cu mae amr
(c) (One remote group member) create and share a Google drive or Box folder with the rest of the group.
Add subfolders for each group member to put their code in.
(d) Put all the required files in the shared folder.
1.2 Station Set-up - In-Person student
(a) Join the lab Zoom session (link on Canvas) on the lab computer and share your screen.
(b) Open the shared folder created by one of the online group members. Add your files.
(c) Create a local folder on the lab computer and copy all the files there.
(d) Open Matlab and change the working directory of Matlab to be the folder that contains your files.
Make sure to periodically save data to your online folder.
(e) Unplug your iRobot Create. Make sure to only pick up the robot from the bottom– not the sensor
platform. Be careful not to hit the markers on the top of the robot. Put it on the floor
next to your lab station.
(f) Take note of the name of your robot.
(g) Turn on the robot by pressing the power button. The Raspberry Pi takes about 20-30 seconds to
boot.
(h) In your MATLAB Command Window, run Robot = CreatePiInit(‘robotName’) where robotName
is the robot name from step (f). The initialization process creates the variable Robot, which contains
the port configurations for interfacing with the Create and the added sensors and the robot name;
it has five elements:
• Robot.Name contains the robot name.
• Robot.OL Client used for getting the robot pose ground truth from the Optitrack system.
• Robot.CreatePort used for sending commands and reading sensors from the robot.
• Robot.DistPort used for getting depth information from the realsense camera.
• Robot.TagPort used for getting tag information from the realsense camera.
(i) Check that you have connected to the robot properly by running BeepRoomba(Robot.CreatePort).
You should hear your robot beep.
(j) Put your robot on the field. Make sure you can access the robot’s localization information by running
[x,y,theta] = OverheadLocalizationCreate(Robot).
(k) Put the robot in front of a tag, at least one foot away. Run RealSenseTag(Robot.TagPort) and
RealSenseDist(Robot.DistPort). Make sure you can get tag and depth information.
(l) If any of steps i-k fail, disconnect from the robot (run CreatePiShutdown(Robot)), shut it down,
close Matlab, restart Matlab, turn the robot on and go back to step h.
Important:
• If the control function exits with an error, make sure to stop the robot by typing in the command
window: SetFwdVelAngVelCreate(Robot.CreatePort, 0, 0)
• When you are done working with the robot, or you wish to restart Matlab or the connection to
the robot, first run CreatePiShutdown(Robot) to disconnect properly from the robot.
• Make sure to alternate the group member that is typing in the Matlab window. Switch
between every section of the lab manual. Remote students: this will be done by remote screen
sharing.
1.3 Robot Motion
In this part of the lab you will collect data and observe how accurately the robot moves.
(a) The function OpenLoopControl.m controls the robot to perform several different motions - first
straight line motions at different speeds, and then rotations at different angular speeds. Place the
robot such that it is aligned with one of the global axes and it has at least 10 feet free in front
of it. Run the function. Make sure your robot has enough space to run the sequence of
motions without running into the walls or other robots.
(b) In the MATLAB Command Window, access the overhead localization data by typing: global
poseData; (or just call the function with poseData as the output). You should have 12 rows.
Save to a file.
(c) Edit the function OpenLoopControl.m to create different motions, for example, you can try
SetFwdVelAngVelCreate for a fixed time. Collect and save the pose data and the commanded
actions.
(d) In the lab report you will be asked to compare the expected motion (based on the function) and the
actual motion (based on the overhead localization). Make sure you have data for both. You will
notice that sometime the overhead localization does not provide data; this will appear as NaN in
the output. Make sure you have enough data to complete the lab report, and if you do not, run the
functions again starting at a different initial pose.
1.4 RealSense Sensor Measurements
The realsense data: In this class, we process the RealSense information to create two types of measurements: depth and beacon (also referred to as tag) relative position. The depth information contains
9 points going from left (the angle is 27 deg in the sensor frame) to right (the angle is -27 deg in the
sensor frame). The depth information is taken from the middle of the camera image.
Processing time of the RGBD information is not negligible. The Raspberry Pi continuously runs
a loop that requests an image from the camera and then calculates depth and tag information. The
information received by the depth and tag functions also contains the time elapsed since the image was
taken.
In this part, you will find the sensor frame location in the robot frame and observe possible error sources
for the depth and tag information.
(a) Choose a beacon (AprilTag).
(b) Position the robot such that the sensor is facing the beacon and the robot center is 2 ft away from
the beacon. It is important that the center of the robot is aligned with the center of the beacon and
the robot X-axis is perpendicular to the beacon plane.
(c) Read the beacon location using RealSenseTag(Robot.TagPort). What are the different elements
of the resulting measurement? What units are the measurements in? (Hint: you can type help
RealSenseTag). You will use the measurements corresponding to the x and y position of the tag.
(d) Calculate the difference between the distance measured on the floor and the distance along the xaxis measured by RealSenseTag(Robot.TagPort). This is the offset in the x-axis. The offset in the
y-axis is the value measured by RealSenseTag(Robot.TagPort) in the y-axis.
(e) Read the distance information using RealSenseDist(Robot.DistPort). What are the different
elements of the resulting measurement? What units are the measurements in? (Hint: you can type
help RealSenseDist). Save the middle measurement (angle 0). Repeat the measurement a few
times and average the distance results. Write down the measurements and the average.
(f) Position the center of the robot 3 ft away from the beacon. Read the beacon location. Repeat the
measurement and calculation. The x-axis offset will be the average of the two x-axis offsets, and the
y-axis offset will be the average of the two y-axis offsets. Make sure to write down the offsets - you
will use them in the lab report.
(g) Read the distance information. Save the middle measurement (angle 0). Repeat the measurement a
few times and average the distance results. Write down the measurements and the average.
(h) Place the robot next to a wall, facing the opposite wall, along the width of the map, such that
the robot X-axes is perpendicular to the wall. Read the distance information. Save the middle
measurement (angle 0). Repeat the measurement a few times and average the distance results.
Write down the measurements and the average.
1.5 Collecting Sensor Data
In this part of the lab you will collect sensor data and observe how delays and sensors data frequency
depend on the functions you use.
(a) Edit turnInPlace.m, set the angular velocity cmdW to 0 and comment out the Pause function (the
robot should remain stationary, and gather data). Place the robot such that it sees a corner and at
least one beacon. Run turnInPlace.m for 30 seconds.
(b) In the MATLAB Command Window, access the sensor data by typing: global dataStore; (or just
call the function with dataStore as the output). Make sure all the sensor data has been properly
recorded (pose, odometry, bump, depth, beacon) – you may want to plot it – and save to a file.
(c) Run turnInPlace.m (with cmdW set to 0 and no pause) for 30 seconds after modifying the functions
as follows. For each run, make sure all the relevant data has been recorded and save to file.
(i) In the function readStoreSensorData.m comment out all sensor readings except for the overhead localization.
(ii) In the function readStoreSensorData.m comment out all sensor readings except for the odometry (both distance and angle).
(iii) In the function readStoreSensorData.m comment out all sensor readings except for the bump.
(iv) In the function readStoreSensorData.m comment out all sensor readings except for the camera
functions – depth and beacon.
(v) In the function readStoreSensorData.m uncomment all sensor readings (i.e. collect all the
data) but in turnInPlace.m comment out the command that sets the velocity for the robot.
(d) Uncomment the command that sets the velocity for the robot. In the function readStoreSensorData.m
comment out all sensor readings except for the camera functions – depth and beacon. Run turnInPlace.m
(with cmdW set to 0 and no pause) for 30 seconds for the following cases. For each run, make sure
all the relevant data has been recorded and save to file. To figure out how many beacons the robot
detects, run RealSenseTag(Robot.TagPort) before running turnInPlace.m.
(i) The robot does not detect any beacons.
(ii) The robot detects one beacon.
(iii) The robot detects at least two beacons.
(e) In the function readStoreSensorData.m uncomment all sensor readings (i.e. to be able to collect
all the data).
1.6 Running Your Control Program: Bump
In this part of the lab you will run your control function backupBump, collect sensor data for analysis
and observe the behavior of the robot.
(a) Open backupBump.m (from any of the group members), and ensure that it calls the function
readStoreSensorData.m to gather and store sensor data.
(b) Using limitCmds.m, set the robot’s maximum velocity to 0.1m/s.
(c) Run backupBump.m and allow the robot to “bump” into the walls a few times before terminating the
program. Make sure you run your program when the robot is in view of beacons.
(d) In the MATLAB Command Window, access the sensor data by typing: global dataStore; (or just
call the function with dataStore as the output). Make sure all the sensor data has been properly
recorded (pose, odometry, bump, depth, beacon) – you may want to plot it – and save to file.
(e) Place the robot close to a wall and run backupBump.m. Observe the behavior.
(f) In the function readStoreSensorData.m comment out all sensor readings except for the bump sensor.
Place the robot close to a wall, run backupBump.m and observe the behavior. Part 2.2 will ask about
the difference in behavior and responsiveness, if any was observed.
(g) Uncomment the other sensors in readStoreSensorData.m.
1.7 Visiting Waypoints
In this part of the lab you will control the robot so that it visits a set of waypoints in the map and
observe the effects of  on the resulting trajectories.
(a) Open your function visitWaypoints.m, and set your waypoints appropriately (coordinate the waypoints with the TA so several groups can run at the same time).
(b) Using limitCmds.m, set the robot’s maximum velocity to 0.1m/s.
(c) Run visitWaypoints.m to have the robot visit the series of waypoints.
(d) Once the robot has followed its path, access the sensor data in the MATLAB Command Window by
typing: global datastore; (or just call the function with dataStore as the output). Make sure
all the sensor data has been properly recorded – you may want to plot it – and save to file.
(e) Repeat for each member of your group. Use the same waypoints, but change the value of  that is used
by the feedback linearization. Make sure to save (or write down) the waypoint coordinates
and the value of  for each run.
1.8 Clean up
When you are done collecting all the data and after you make sure you have everything that you need
for the lab report:
(a) Run CreatePiShutdown(Robot) to disconnect from the robot.
(b) Turn off the robot, return it to the lab station and plug it in.
(c) Make sure to leave your lab station clean, without any papers or other items. Follow COVID-19
protocols for sanitizing your station.
2 Post-Lab Assignment
The group should submit one post-lab report Lab1report.pdf, on Canvas, one week after the lab took
place, by 11:59 PM.
2.1 Actuation and sensor errors (20 points)
(a) In 1.3, the function OpenLoopControl.m commands the robot to perform a series of motions. Create
two figures, one for the distance traveled and one for the angle turned. On each of these figures,
plot the actual data (from your saved PoseData) and the expected motions based on the function.
Plot each motion segment (one command in OpenLoopControl.m, either a forward motion or angle
turned) in a different color and use a different line style to distinguish between the expected and the
actual motions.
(b) Based on the figures and analyzing the difference between the expected and actual motions, comment
on the actuation errors you observe. Does the robot perform the desired motion perfectly? does the
error, if any, depend on the speed? How can we mitigate such errors? How fast do you think we
should drive the robot?
(c) For the depth data collected in 1.4, taking into account the location of the sensor with respect to the
robot (estimated in 1.4), calculate the difference between the actual measurements and the expected
measurements. Do they match? is the error constant? does it change with the distance? what could
be the sources of the error? comment on the precision and accuracy of the depth sensor.
(d) If the distance measurement error is a constant bias (a constant error regardless of the distance),
how would you take that into account when using the measurements? if the error is not constant,
how would you take that into account?
2.2 Sensor frequencies and delays (25 points)
(a) For the data sets obtained in 1.5(a)–(c) observe and compare the timestamps of the data. What do
you notice? how often can you get the different sensor information? how does this frequency depend
on your code?
(b) What are the implications of these delays on your control code and the robot behavior? how might
you mitigate any unwanted behaviors?
(c) For the dataset in 1.5(a) what do you observe about the timing of the depth and tag information
with respect to the other sensors? When doing localization (as you will do in Lab 2), how should
you deal with the delay in these measurements?
(d) For the datasets in 1.5(d), how does the number of tags affect the delays?
(e) In Part 1.6(e)–(f), did the robot react differently to the presence of a wall? How so?
(f) If so, what is the reason for the different behavior? If the behavior was identical, why would one
expect commenting out parts of the code would lead to different behaviors?
2.3 Sensor Information (35 points)
(a) Plot the robot trajectory from Part 1.6(a)–(d).
(b) Indicate places where the robot’s bump sensors were triggered (for example, with a *).
(c) How might you build an obstacle map if you only had information from the robot’s bump sensors?
(d) Using the function robot2global.m, and the location of the camera in the robot-fixed frame you
obtained during calibration, plot the locations of any detected beacons (indicate their locations with
a different symbol than was used in part (b).) Keep in mind that the robot pose timestamps are
not perfectly aligned with the beacon timestamps. How did you handle the timing mismatch? Did
you pick the “closest” timestamp? Did you interpolate the data, and if so, which data? Justify your
choice.
(e) Discuss the precision of the camera. Are the locations of the detected beacons consistent, or do they
move around a lot?
(f) Write a function depth2xy.m to convert the raw depth measurements into 2D local robot coordinates
(this function will be similar to lidar range2xy.m from Homework 1.) Make sure to take into
account the location of the camera you obtained during the lab. The minimum distance for the
depth sensor is 17.5cm; if an obstacle is closer, the sensor will return 0. How are you taking into
account the nonlinearity in the range error, if at all?
(g) Plot the lab map and the robot’s depth data from Part 1.6(a)–(d) in global coordinates (you may
use the function robot2global.m from Homework 1 for the depth data). Keep in mind that the
robot pose timestamps are not perfectly aligned with the depth timestamps. How did you handle
the timing mismatch? Did you pick the “closest” timestamp? Did you interpolate the data? Justify
your choice.
(h) How well did the depth sensor measure the truth map? For example, are the walls detected as
straight lines? What may be some sources for errors?
(i) Load the depth returns from the stationary robot run in Part 1.5(a)–(c) (all the runs that have depth
information). If the depth noise is modeled as a Gaussian, wdepth ∼ N
0, σ2


, estimate σ.
2.4 Waypoint Follower (25 points)
(a) Plot all the robot trajectories from Part 1.7 on the same figure (you should have one for each member
of the group) and indicate which  was used by each robot for feedback linearization.
(b) Plot the waypoints on the same figure.
(c) Comment on any differences between the robots’ trajectories. Why wouldn’t the trajectories be
exactly the same? Did members of the group handle this task differently?
(d) How did the value of  for feedback linearization affect the robots’ trajectories? How would the
trajectory have been different for a holonomic vehicle?
(e) For one of the  values, run the same waypoints and the same code in the simulator. Plot the
resulting (simulated) trajectory. On the same figure, plot the actual trajectory. Do you expect both
trajectories to be the same? Are they? If they are not the same, explain what the possible sources
of difference are.
2.5 Lab preparation (mandatory, not graded)
(a) Which robot did you use?
(b) For the code you wrote as part of Homework 1, were there any coding errors that were not discovered
when using the simulator?
(c) How much time did you spend debugging your code in the lab?
(d) Did you observe any behaviors in the lab that you did not expect, or that did not match the simulated
behaviors?
(e) What was the contribution of each group member to the lab and the report