S25 Micro Quests

Micro Autonomy Quest Book - Spring 2025

The Great Objective: Fully Autonomous F1tenth Racing Car with LiDAR and Camera

Micro Autonomy aims to win the F1teneth Autonomous Racing Competition in the future

Term Objectives and Tasks

  • 2.3. Camera Integration with Localization: Integrate Camera Object Detection with Localization to use real-time slam mapping
  • 5. Planning: Local Planning: Enable Obstacle Avoidance in lattice planner on actual vehicle
  • 6. Control: Pure Pursuit: Integrate and tune algorithm with local planner to work on actual vehicle

Hardware Setup

1.1 Traxis Car Assembly

ScoreCriteria
10/10Fully built with all sensors and hardware mounted and power system working
8/10Have all the sensors and Jetson mounted to the Platfrom Plate
6/10Fabriate the Platfrom mounting plate and all other mounting hardware
4/10Gut and stip all the components from the original chassis
2/10Purchase & manufacture all (within current budget) components
0/10No Progress

Progress: 10/10

1.2 LiDAR

ScoreCriteria
10/10Setup interfaces for LiDAR to ouput a ROS /scan topic
7/10setup the docker compose file to correctly interface with the LiDAR
5/10configure the ip and network settings for the LiDAR
0/10No Progress

Progress: 10/10

1.3 Vec

ScoreCriteria
10/10tune the IMU to correctly how roll, pitch and yaw data and accleration values
7/10tune the motor PID controller to produce a step response
5/10configure all the hardware limits and current settings
2/10Have the vesc power on and showup in the Vesc tool software
0/10No Progress

Progress: 10/10

1.4 Overall Hardware integration

ScoreCriteria
10/10Setup interface with the correct ROS Drivers and have a tuned and accurate odometry output
7/10configure the yaml files to fine tune odometry
5/10Install the correct ROS Transport drivers along with F1teneth Driver Stack
0/10No Progress

Progress: 10/10

State Estimation and Localization

2.1 Extended Kalman Filter

ScoreCriteria
10/10Have a fully functional Extended Kalman Filter working on the physical vehicle
8/10Have a fully functional Extended Kalman Filter working in the simulation
6/10Have fully Defined sensor models for the EKF
4/10Have Fully Defined motion model for the EKF along with the corresponding Jacobian
2/10have fully Defined state to propigate for the EKF
0/10No Progress

Progress: 10/10

2.2 Particle Filter

ScoreCriteria
5/5Have Particle Filter working on physical vehicle
3/5Have Particle Filter working in the simulation
0/5No Progress

Progress: 10/10

2.3 Integrate Camera Object Detection with Localization

ScoreCriteria
10/10Localization works for changing environments in actual vehicle
5/10Ignores LiDAR scans that interfering with real-time mapping
0/10No Progress

Progress: 0/10

Mapping: SLAM

ScoreCriteria
10/10Map and yaml file generated through state estimation and LiDAR scans on actual vehicle
8/10Map and yaml file generated through state estimation and LiDAR scanes in simulation
5/10Map and yaml file generated through true position (odometry) and LiDAR scans in simulation
0/10No Progress

Progress: 10/10

Raceline Generation/Optimization

ScoreCriteria
10/10Optimizes centerline to generate a raceline and velocity profile that minimizes steering for all maps
7/10Optimizes centerline to generate a raceline that minimizes steering for all maps
5/10Generates a centerline for the vehicle to follow
0/10No Progress

Progress: 10/10

Planning: Lattice Planner

ScoreCriteria
10/10Genenerates/optimizes local trajectory based on cost map and vehicle dynamics to achieve obstalce avoidance on actual vehicle
9/10Genenerates/optimizes local trajectory based on cost map and vehicle dynamics to achieve obstalce avoidance
8/10Generates local trajectory by using a cost map to achieve obstacle avoidance when obstalces are on the ideal path
5/10Generates local trajectory to make vehicle follow the raceline (ideal path)
0/10No Progress

Progress: 5/10

Controls: Pure Pursuit

ScoreCriteria
10/10Follows local trajectories and ideal velocities smoothly on actual vehicle
7/10Follows local trajectories and ideal velocities smoothly in simulation
4/10Follows a global trajectory by controlling steerring angle with a constant velocity
0/10No Progress

Progress: 7/10

Camera Object Detection and Tracking

ScoreCriteria
10/10Camera detects opponent cars and returns bounding boxes reliabilty in real-time with low latency
6/10Camera detects opponent cars and returns bounding boxes
0/10No Progress

Progress: 0/10

Scoring Template Summary

Quest NameDescriptionDue DateScore
Hardware Setup 1.1Fully built with all sensors and hardware mounted and power system working2025-08-3110/10
Hardware Setup 1.2Setup interfaces for LiDAR to ouput a ROS /scan topic2025-08-3110/10
Hardware Setup 1.3Tune the IMU to correctly how roll, pitch and yaw data and accleration values2025-08-3110/10
Hardware Setup 1.4Setup interface with the correct ROS Drivers and have a tuned and accurate odometry output2025-08-3110/10
State Estimation 2.1Have a fully functional Extended Kalman Filter working on the physical vehicle2025-08-3110/10
State Estimation 2.2Have Particle Filter working on physical vehicle2025-08-315/5
State Estimation 2.3Integrate Camera Object Detection with Localization to use real-time slam mapping0/10
Mapping: SLAMGenerate a map in image format and a yaml file for the map specifications based on IMU, encoder and LiDAR specifications.2025-08-3110/10
Raceline Generation/OptimizationGenerate a optimized raceline as global trajectory (ideal path) for the race car to follow to minimize lap time.2025-08-3110/10
Planning: Lattice PlannerGenerate local trajectories for obstacle avoidance and following global trajectory during Racing.2025-08-315/10
Controls: Pure PursuitControls the vehicle's steering and trottle to following the trajectories generated by the planning module.2025-08-317/10
Camera Object Detection and Edge TrackingCamera detects opponent cars and returns bounding boxes reliabilty in real-time with low latency0/10