How to build Maker's Pet Mini (LiDAR ROS 2 education robot): the whole process
206 steps to build Maker's Pet Mini (LiDAR ROS 2 education robot), from buying parts, printing and assembly to software, calibration and training, with the bill of mater…
120mm Build Pack $77–150 (Maker's Pet official store, 2026‑11‑01, USD): $150 with printed parts and the LD14P LiDAR, down to $77 if you print the parts and bring your own LiDAR; batteries not included, sales tax only on orders shipped to California, free worldwide shipping over $150 per the home page, and import taxes on international orders paid by the buyer. Parts sold separately: BDC‑30P driver board $35, LD14P LiDAR $59, 120mm body from $20, N20 encoder motor $10, ESP32E dev board $9
Buy parts
- Prepare the computer and WiFi
Use a local PC running Windows with at least 8GB RAM or Linux with at least 4GB RAM, and connect to a 2.4GHz WiFi network.
- Choose and order the selected parts
Choose the parts according to your selected options. The full-option order includes the LD14P LiDAR and cable, N20 encoder motors and accessories, wheels, battery hardware, fasteners, base, caster, ESP32 board, and BDC-30P motor driver.
- Read the rechargeable battery warnings
Before using the rechargeable battery option, read the additional instructions in the WARNINGS section.
- Prepare a 2.4 GHz network
Join the robot and PC to the same 2.4 GHz WiFi network. Do not use a 5 GHz network because the ESP32 cannot use it.
- Select a USB data cable
Use a USB data cable for the ESP32. A cable that only carries power will not work.
- Gather assembly tools
Gather a small screwdriver and a wire stripper before assembling the robot.
Print and make
- Download support and caster parts
Download the 3MF/STL files for lidar skirts, lidar posts, board posts, caster roller, caster mounts, and the 2S battery holder backstop.
- Download the robot base
Download the 3MF/STL file for the robot base from the v1.0.1 download.
- Download the wheels
Download the 3MF/STL files for the wheels.
- Download the motor clamps
Download the 3MF/STL files for the motor clamps.
- Print the 3D-printable robot parts
Print all 3D-printable parts yourself in your preferred color. When doing this, set the “3D Printed Parts” option to “None” to save some money.
Assemble
- Strip the motor wire
Strip 8–10 mm of insulation from a motor wire.
- Fold back the bare strands
Bend the bare strands back over the insulation; the thicker end grips better.
- Insert the motor wire
Insert the wire with the bare strands facing up. If using a metal contact, place the metal on top.
- Tighten the terminal screw
Tighten the terminal screw to secure the inserted motor wire.
- Connect both motor cables
On both terminal blocks, use M2 red, M1 white, GND blue, ENCA green, ENCB yellow, and +Venc black. If cable colors differ, follow the plug pin order and board terminal labels, then tug each wire gently.
- Attach the four board posts
Attach the four shorter board posts to the base plate with countersunk M3 screws. Twist each post and its tab carefully into the shown orientation.
- Attach the four LiDAR posts
Attach the four longer LiDAR posts to the base plate with countersunk M3 screws and orient them as shown.
- Attach the caster mounts
Attach the two caster wheel mounts to the back of the base with hex button M3 screws, leaving them slightly loose.
- Insert the caster shaft
Push the metal shaft through the caster roller.
- Seat the caster roller
Place the two shaft ends into the caster mounts.
- Tighten the caster mounts
Tighten the caster mount screws to secure the roller assembly.
- Free the caster roller
Press the roller against the table and roll the base back and forth until the roller spins freely.
- Insert the motors into clamps
Insert each motor into its clamp with its front face flush against the clamp.
- Mount both motors
Attach each motor to the base with three countersunk M3 screws and point the shafts outward. Do not overtighten, because the plastic base can bend.
- Check the Board Post Tabs
Confirm that no board post tab touches a motor. If a tab presses against a motor and blocks it, twist the post to turn the tab away.
- Tighten the Motor Screws
Check that bent plastic, including the base, cannot rub against the motor gears or encoder disk. Tighten the motor screws just until snug.
- Clear Debris From the Wheels
Make sure each wheel is free of debris before installing it.
- Seat the Tires in the Grooves
Press each tire into its wheel's groove all the way around using your fingers.
- Locate the Shaft and Hub Flats
Find the flat on the motor shaft and the matching flat in the wheel's hub.
- Press the Wheels Onto the Shafts
Line up the motor-shaft flat with the matching wheel-hub flat. Press the wheel fully onto the shaft with both hands.
- Clean the Wheels and Hubs
Clean the wheels and hubs so that grit or plastic shavings cannot jam the gearbox gears and burn out the motor.
- Connect the Motor Cables
Align each motor cable plug with the connector on the back of its motor. Press the plug in gently; if it will not enter, check its alignment instead of forcing it.
- Route the Motor Cables
Route the motor cables underneath the board. Gently bend the wires downward at the motor plugs.
- Attach the Board to the Posts
Attach the board to the posts using four hex button M3 screws.
- Inspect the Motor Connector Housings
Before plugging in each cable, check the motor connector because its plastic housing can slide up off its pins.
- Connect the Battery Wires
Connect the battery case wires to the +BAT, GND, and ExtVmot screw terminal. Connect red to +BAT and black to GND.
- Check the Battery Polarity
Check the battery polarity before switching on. Reversed polarity will burn the board.
- Insert the Conductors Straight
Do not fold back the stripped conductors. Insert them straight into the screw terminal.
- Orient the ESP32 Module
Orient the Maker's Pet ESP32-E 30-pin dev kit module with its USB connector toward the battery screw terminals.
- Check the ESP32 Power Pins
Check the module orientation before pressing it fully into the socket. A backwards module gets the wrong power pins.
- Insert the ESP32 Module
Press the correctly oriented ESP32-E module fully into the board's socket.
- Follow the assembly video
Follow the follow-along video instructions to assemble the robot.
- Use the free e-book
Alternatively, follow the instructions in the free e-book.
- Read rechargeable battery warnings
Read the instructions and rechargeable (protected only!) battery use warnings in the FAQ section. Using rechargeable batteries to power the robot is AT YOUR OWN RISK!
- Mount rear posts
Mount the rear posts using the outer spare holes in the base.
- Install battery holder backstop
Mount the battery holder backstop behind the battery, using two M3 countersunk screws in the two spare holes.
- Mount LiDAR with wider skirt
Mount the LiDAR using a wider skirt that matches the positions of the rear posts.
- Keep PCB Tab Clear
Make sure the PCB post tab does not touch the motor. Blocking motor rotation in any way can burn the motor when it is powered up.
- Seat Motor Connector Housing
The motor connector housing can slide off toward the motor’s magnet. Pull the connector housing carefully back into place.
- Check Motor Terminal Wiring
Make sure the motor wires are connected to the correct board screw terminals.
Software
- Download robot firmware
Download the robot firmware.
- Install SPIFFS plugin for Arduino IDE 2.x
Install the SPIFFS Upload plugin for Arduino IDE 2.x.
- Install SPIFFS plugin for Arduino IDE 1.8.19
Install the SPIFFS Upload plugin for Arduino IDE 1.8.19.
- Install Microsoft WSL
Install Microsoft WSL using the linked installation instructions.
- Install Docker for Windows
Install Docker for Windows.
- Install X server and set display
Install the X server and set the Display number to 0 (zero).
- Install PowerShell
Install PowerShell.
- Install ROS2 Iron and Kaia.ai
To skip Docker, install ROS2 Iron and the Kaia.ai packages directly on Ubuntu 22.04 using install_ros2_iron_ubuntu_22_04.sh and then install_kaiaai_iron.sh.
- Use the Iron Docker image
Use the kaiaai/kaiaai:iron Docker image for these instructions.
- Start the Jazzy Docker image
Start the kaiaai/kaiaai:jazzy image on Linux with start_jazzy.sh.
- Enable shared-memory messaging
When writing a custom docker run command with --net=host, also add --ipc=host so external ROS2 nodes receive messages.
- Install Ubuntu in a virtual machine
Use this approach only if installing WSL2 is troublesome. Install Ubuntu 22.02 on the Windows PC with a virtual machine instead of WSL2.
- Run the KaiaAI Docker container
Run the Docker container with the shown port mappings. If Docker reports that TCP port 4430 is unavailable, reboot the Windows PC.
- Stop the existing makerspet container
If Docker says that the makerspet container name is already in use, stop the existing container and retry the docker run command.
- Pull the KaiaAI Iron image
Pull the kaiaai/kaiaai:iron image with Docker.
- Launch Docker Desktop
If Docker reports that the Docker engine is not running, launch Docker Desktop on the Windows PC.
- Preserve container changes
Changes made inside the running makerspet container are lost when a fresh container is started. Commit the running container to the image to keep the changes, or push them to your own Docker Hub repository as an off-PC backup.
- Back Up the Iron Image
Optionally tag the existing kaiaai Iron image for your own Docker Hub repository and push it.
- Install Ubuntu on the Raspberry Pi
Install Ubuntu 22.04 natively on the Raspberry Pi because the Docker image supports x86-64 rather than ARM64.
- Install ROS2 Iron Natively
Install ROS2 Iron on Ubuntu 22.04 with install_ros2_iron_ubuntu_22_04.sh, or use the official ROS2 Iron Ubuntu .deb instructions as a fallback.
- Install the Kaia.ai Add-ons
Install the Kaia.ai add-ons with install_kaiaai_iron.sh after installing ROS2 Iron.
- Install ROS2 Jazzy Natively
For ROS2 Jazzy, use Ubuntu 24.04 and install ROS2 with install_ros2_jazzy_ubuntu_24_04.sh.
- Install the Jazzy Add-ons
Install the Kaia.ai add-ons for ROS2 Jazzy with install_kaiaai_jazzy.sh.
- Run Several Robots in Docker
Run one Docker container per robot with a unique container name and a different host UDP port. Point each robot’s WiFi configuration at its matching port.
- Enable True Topic Namespacing
Edit the YAML files in /ros_ws/src/makerspet_mini/config, such as telem.yaml, to apply Turtlebot3-style true topic namespacing.
- Omit IPC Host for Jazzy
Containers started with the Jazzy image each receive their own network, so do not add --ipc=host.
- Start the X Server First
Make sure the X server is running before launching RViz or any other GUI program under Windows WSL.
- Stop Stuck ROS2 Processes
Open another Bash window, use ps -al to identify the ros2 and python3 processes, then stop both processes.
- Launch Windows X Server
Launch the X server on the Windows PC before using ROS2 graphical interfaces.
- Use VNC for ROS2 GUIs
TigerVNC and Xfce are bundled in the Docker image. Start the container with the VNC port and display set, start VNC inside it, and connect a VNC viewer to localhost:5901.
- Tune Third-Party Motor PID
For a third-party motor not purchased from Maker’s Pet, tune the motor’s PID coefficients in the driver board’s config.yaml file if needed. Carefully solder any loose motor encoder components.
- Configure a Supported Motor Driver
The firmware controls the motor through the IN1/IN2 inputs of TB6612FNG, DRV8871, DRV8833, DRV8835, or L298N drivers. Hold the driver enable pins HIGH: tie PWMA, PWMB, and STBY high for TB6612FNG, or tie ENA and ENB high for L298N through its jumpers or to 5V.
- Map Driver Type and GPIOs
In config.yaml, set IN1_IN2 for the supported drivers or PWM_CW for a BLDC motor with a built-in controller, and map each motor’s GPIOs. If a wheel spins the wrong way or will not stop, swap that motor’s in1/in2 connections or encoder a/b connections; the document links to a TB6612FNG ESP32 worked example.
- Prevent Startup Motor Spinning
Some ESP32 motor-driver GPIOs float or go HIGH at boot or reset, causing a motor to spin at full speed before firmware control begins. At the start of setup(), explicitly drive the motor GPIO pins LOW.
- Configure the BLDC motors
Set motor.driver.type to PWM_CW and motor.encoder.type to FG in config.yaml, then map the FG, PWM, and CW GPIOs for each side.
- Select the ESP32 board
In Arduino IDE, go to Tools -> Board -> ESP32 Arduino and select the ESP32 board, usually ESP32 Dev Board.
- Choose a compatible ESP32 package
Do not use ESP32 board package 2.0.15 because HardwareSerial is broken. Use 2.0.16 or newer, with 2.0.17 recommended; stay on 2.0.17 because ESP32 core 3.x removed ledcAttachPin and ledcSetup.
- Prevent bundled library replacement
Rename the kaiaai libraries, for example to MotorController_kaia, so Arduino IDE 2.x’s Library Manager does not silently replace them with same-named libraries by other authors.
- Upload the ESP32 sketch data
Upload sketch data for either the 30-pin ESP32 dev kit or the 38-pin ESP32 dev kit with the ESP32-WROOM-32D module. After upload, the serial monitor should show that SPIFFS mounted successfully.
- Enter the ESP32 web configuration
Push and release EN, then push and hold BOOT within one second to enter web configuration.
Calibrate and test
- Turn On Battery Power
Make sure the battery power switch is ON.
- Check ESP32 Power LED
With the battery power switch ON, check that the power LED on the ESP32 board lights up.
- Check Motor Encoder LED
Check the green LED on the motor. If it is OFF, the motor’s encoder is not receiving power, which can result from a disconnected battery, an off power switch, a failed motor-to-cable connection, or a failed cable-to-board connection.
- Free the Encoder Magnet
Carefully bend the encoder sensors back to their normal position. Confirm that the encoder magnet rotates freely without touching the sensors.
- Test Forward Motor Rotation
Run the ROS2 teleoperation app and press the w key. Confirm that the robot’s motor rotates forward.
- Check an Unmoving Motor
Check for cables blocking the encoder magnet, mechanical stalling of either motor, reliable M1 and M2 board connections, and fresh batteries. Swap the motors to isolate the fault; the assembly video shows a case where the metal conductors failed to make contact.
- Correct One-Way Full-Speed Rotation
For a motor that stays at full speed in one direction regardless of the command, swap its ENCA and ENCB wires. Verify that both encoder wires are reliably connected to the board and motor and inspect any bad connection.
- Correct Reversed Motor Rotation
If a motor runs in the opposite direction but otherwise responds correctly, swap its ENCA and ENCB encoder wires, then swap its M1 and M2 wires. Check for flaky wire connections.
- Use the Correct 12V Battery
For 12V Vmot, close JP2 as the default and use Vbat of about 6–10V. A 3S LiPo at 11.1V is not reliable; use a 2S pack at approximately 7.4V.
- Use the Correct 24V Battery
For 24V Vmot, open JP2 and use a Vbat of approximately 6–21V.
- Use sufficient encoder pulse resolution
Use roughly 1000 PPR or higher; a very low-PPR FG signal can make low-speed PID control twitch.
- Read the ESP32 status LED
Solid ON means AP mode; slow blinking once per second means connecting to WiFi; very slow blinking once per 20 seconds means connected to WiFi while connecting to the ROS2 PC; fast blinking means connected to the ROS2 PC.
- Verify the loaded robot configuration
Confirm that /network.yaml and /config.yaml are loaded, the board is MINI-BDC30P-BODY with BDC-30P version v1.1.1, and the LIDAR is LDROBOT LD14P with a 1024-byte RX buffer and 230400 baud rate. The reported battery ADC attenuation is 7.00 with voltage 2.60V, and the motor driver type is IN1_IN2 with AB_QUAD motor encoder type.
- Verify board and motor configuration
Confirm the reported board model, motor-driver type IN1_IN2, encoder type AB_QUAD, maximum motor speed of 180.00 RPM, encoder PPR of 1035.00, and TPR of 4140.00.
- Enter the web configuration
Push and release EN, then push and hold BOOT within 1 second to enter web configuration.
- Verify the LiDAR configuration
Confirm that the LiDAR is the LDROBOT LD14P, uses a 1024-byte RX buffer and 230400 baud rate, and that startLIDAR() reports OK.
- Increase LiDAR Spin-Up Threshold
If a motor-RPM error occurs before the LiDAR reaches speed, raise scan_rpm_err_thresh to give it more time to spin up.
- Choose a Docker Desktop backend
Choose WSL 2, Hyper-V, or Docker VMM. WSL 2 is the default and works for most users without administrator privileges.
- Verify the WSL version
Use WSL version 2.1.5 or later. If using Enhanced Container Isolation, use WSL version 2.6 or later.
- Check the Windows version
Use Windows 10 version 22H2 build 19045, or Windows 11 version 23H2 build 22631 or higher, with Enterprise, Pro, or Education edition.
- Enable the Windows Server service
Enable the Windows Server service (LanmanServer) and set its start mode to Automatic.
- Check Hyper-V Windows requirements
For Hyper-V, use Windows 10 version 22H2 build 19045, or Windows 11 version 23H2 build 22631 or higher, with Enterprise, Pro, or Education edition.
- Enable the service for Hyper-V
Enable the Windows Server service (LanmanServer) and set its start mode to Automatic.
- Verify processor, memory, and virtualization
For Client Hyper-V on Windows 10, use a 64-bit processor with SLAT and 8GB system RAM. Turn on hardware virtualization support in the BIOS/UEFI settings.
- Check Windows version support
Check that your Windows version is still within Microsoft’s servicing timeline. Docker Desktop is not supported on server versions of Windows, such as Windows Server 2019 or Windows Server 2022.
- Confirm the Windows edition
Running Windows containers requires Windows 10 or Windows 11 Professional or Enterprise edition. Windows Home and Education editions can run Linux containers only.
- Configure host virtualization support
Enable hardware virtualization in BIOS/UEFI. When Docker Desktop runs inside a virtual machine, expose nested virtualization from the host.
- Use WSL 2 on Arm systems
Docker VMM is not supported on Arm, and the installer uses WSL 2 instead. Docker VMM does not require the Hyper-V feature.
- Verify the WSL version
Check that WSL version 2.1.5 or later is installed.
- Check the Windows version and build
Use Windows 10 64-bit Enterprise, Pro, or Education version 22H2 (build 19045), or Windows 11 64-bit Enterprise, Pro, or Education version 23H2 (build 22631) or higher.
- Verify WSL 2 hardware requirements
For WSL 2 on Windows 10 or Windows 11, use a 64-bit processor with SLAT and 8GB system RAM. Enable hardware virtualization in BIOS/UEFI.
- Reinstall to change installation mode
To switch the installation mode later, uninstall Docker Desktop and reinstall it.
- Select the container daemon
Use Windows containers only with an all-users Docker Desktop installation. From the Docker Desktop menu, select **Switch to Windows containers** or **Switch to Linux containers**; Linux containers are the default.
- Review configuration for non-default hardware
Read the Kaia.ai firmware config.yaml configuration reference when using non-default motors, LiDAR model, wheels, robot body size, or similar settings.
- Check the Wheel Rotation
Spin each wheel by hand and confirm that it rotates without rubbing.
Data and training
- Monitor robot telemetry
Monitor telemetry for LiDAR RPM, both wheel RPM values, battery voltage, and RSSI to check the robot's operating status.
Use it
- Choose the simulation path
If you do not have the robot, use the PC-based simulation in Chapter 6 to drive the robot, map a room, and send it to goals. The Chapter 7 programs also run on the simulated robot.
Parts
- LDROBOT LD14P 2D LiDAR with a 10 cm breakout cable (optional; other models the firmware supports also work)
- 2 × 12V N20 encoder gear motors (200 RPM), motor breakout cables and a pair of precision motor clamps
- BDC‑30P driver board (53.2×56.4 mm, 31 g): on-board 12V/24V motor supply, two brushed DC motor drivers, two encoder inputs, a LiDAR header with motor PWM, 5V regulator, battery voltage sensing and a power switch, carrying an ESP32 DOIT DevKit v1-compatible 30-pin ESP32/ESP32E dev board
- A 6×AA battery holder, or a 2S 18650 holder with a backstop (batteries not included; the board is designed for non-rechargeable alkaline cells, and 18650s must be protected cells used at your own risk)
- 120 mm body (printed in PETG, or print it yourself): base, 4 LiDAR posts, 4 board posts, LiDAR skirt, caster roller and mounts, two 43 mm wheels, M3 countersunk and button-head screws and an Allen key
Tools
- A computer: Windows with 8 GB+ RAM (WSL2, Docker Desktop and an X server) or Linux with 4 GB+ RAM
- Arduino IDE (ESP32 board package 2.0.16 or newer, 2.0.17 recommended) with the SPIFFS upload plugin, and a data-capable USB cable
- 2.4 GHz Wi‑Fi that lets local devices reach each other (5 GHz is not supported)
- Optional: a 3D printer (printing the body yourself saves money)
Skills
- Mechanical assembly and wiring: motor, encoder and battery leads into the driver board's screw terminals
- Building and uploading the ESP32 firmware in Arduino IDE, uploading config.yaml via SPIFFS, and web-based Wi‑Fi setup
- Running ROS 2 in Docker or Ubuntu: launch files, RViz and keyboard teleop, with Cartographer mapping, Nav2 navigation and explore_lite exploration
- Using non-kit motors, wheels or LiDAR means editing config.yaml and re-tuning the PID (the store advises beginners against it)
Software
- kaiaai/firmware: ESP32 Arduino + micro-ROS firmware with motor PID, encoder and LiDAR drivers, configured per board and parts by config.yaml; latest v0.8.6 (2025‑12‑13)
- kaiaai (ROS 2 package and the kaia CLI): one workspace for several robot models; kaiaai_bringup for mapping, navigation and frontier exploration, kaiaai_gazebo for simulation, kaiaai_teleop for keyboard driving
- Docker images kaiaai/kaiaai:iron and :jazzy (x86‑64); on Ubuntu 22.04/24.04 or a Raspberry Pi, scripts install ROS 2 Iron or Jazzy natively
- makerspet_mini: the Mini's URDF with Nav2 and Cartographer configs