PicoBot with Circular Array of Distance Sensors

• This project is based on its predecessor: PicoBot-oto.
• Although it has been my habit to build DIY projects with whatever components I can find laying around my shop, I found it helpful to build a CAD model for this robot for a couple of reasons:
  • I needed a couple of 3D printed parts for the sensor array.
  • I wanted to make sure the sensor array ended up at the right height to match the lower shelf of the furniture in my living room.

IMG_4341.MOV

Circular Array of Distance Sensors

• 7 VL53L0x time-of-flight VCSEL devices arranged in a 180-deg arc
• Positioned at 9:00 through 3:00 (clock face positions)
  • Sensor A at 9:00 looks straight LEFT
  • Sensor D at 12:00 looks straight ahead
  • Sensor G at 3:00 looks straight RIGHT
• Sensors are connected through an I2C MUX to Pico
• Array data is a list of 7 distance (mm) values [A, B, C, D, E, F, G]

Bi-directional data exchange between laptop and PicoBot

• The Adafruit Bluefruit LE UART Friend device onboard the PicoBot operates as a peripheral device (server) in BLE (Bluetooth Low Energy) communication.
• When the PicoBot is powered up, the BLE UART Friend device begins advertising its presence and waiting for a central device to connect and establish a connection.
• A Python program on the laptop can then be started. It acts as a client, first scanning for the BLE UART Friend, then connecting to it.
• Once connected, bi-directional communications can begin.
  • The client program can send driving instructions to the server.
  • The server can respond by sending messages back to the client.

The example below shows this process in detail.

• Initially, the PicoBot is placed in its home position at coordinates (0, 0), carefully aligned to point exactly in the X direction.
• The PicoBot is switched on, and is undisturbed while the Optical Tracking Sensor self calibrates.
• Next, the controller_display program on the laptop is started.
  • Once the connection with the PicoBot is made, a window opens, displaying a map of the arena boundaries.

• The window has 4 buttons, labeled Drive, Load, Run, and Stop.
• Click the Run button first to start the robot main() function running.
• Next, click the Load button.
  • This causes a generator to be created from a list of driving instructions.
• Finally, click on the Drive button, which gets the next instruction from the generator, sends it to the robot, waits for it to be complete, then gets the next instruction and the next... until there are no more instructions.
  • Each instruction is a one item dictionary, where the key is the drive command and the value is the parameter value.
  • The table below shows the available driving instructions

Instruction	key	value
Send WayPoints to robot	"SWP"	list of points (x, y)
Drive Waypoints (driving Forward)	"DWF"	None (drive the list of points in sequence)
Drive Waypoints (driving in Reverse)	"DWR"	None (drive the list of points in sequence)
Turn in place to Goal Heading	"TGH"	Heading angle (radians)
Turn in place by Relative Angle	"TRA"	Relative angle (radians)

Typical Loop Run

• Here are the driving instructions of a run in which the robot is instructed to drive in a loop around the perimeter of the arena.

wapo_list = [
    (1.0, -0.5),
    (1.9, -1.5),
    (2.0, -2.7),
    (3.1, -3.8),
    (3.2, -5.3),
    (2.4, -6.0),
    (1.5, -6.0),
    (0.9, -5.4),
    (0.8, -4.6),
    (0.5, -4.0),
    (-1.8, -3.8),
    (-4.7, -2.7),
    (-5.6, -1.5),
    (-5.6, 2.0),
    (-4.6, 3.0),
    (-4.0, 3.0),
    (-1.3, 1.4),
    (-0.7, 0.2),
    (0.0, 0.0)
    ]

instrux_list = [
    {"!SWP": wapo_list,},
    {"!DWF": None,},
    {"!TGH": 0,},
    ]

• The instruction list has just 3 lines:
  • Send the waypoint list to the robot
  • Drive the waypoints in sequence
  • Turn in place to Global Heading 0
• During execution of the first instruction, the waypoints are displayed (in Magenta color)
• When the robot begins to execute the second instruction (begins driving to the 1st waypoint), its location is shown as a Blue dot.

• As the robot drives, data points are continually added to the display
  • Robot path is shown as Blue dots.
  • Obstacles detected by the sensors are shown in various colors (different color for each sensor)
    • Reds on the left, Greens on the right, darkest shades abeam, Yellow straight ahead.

• Upon completion of the run, the output of the controller_display program has been manually saved as run_data.py which can then be analyzed by run_analyzer.py.

Total number of lines in file:  3393
Total number of time steps: 570
Average value of time step: 0.16 seconds
Data rate: 6.13 Hz
Duration of run: 93.02 seconds

• Also, at the end of the run, the controller_display program automatically saves all the collected robot data in the file saved_data.py
• The program my_ogm.py constructs an Occupancy Grid Map (OGM) from this saved data.