[CRP] [test] E2E vehicle test - mid RI

[gfigneczi1]

Test date: 10/October/2024

Overall goals: 1) record sample mcaps for simulation later, incl. the /cai/... messages 2) test lane follow step 1, with PID and LDM 3) Build Foxglove UI

Test plan:

• highway scenario: continuous driving in lane with different speeds (50, 90, 120 kph) incl. last curve section, where deceleration happens
• highway scenario: continuous driving, slowing down where the map contains speed limit data
• highway scenario: entry situation with accelerating and lane change, exit situation with lane change and deceleration
• highway scenario: stop&go, traffic jam simulation
• highway scenario: lane change with different speeds
• highway scenario: roadwork simulation
• highway scenario: follow object situation
• handling scenario: test the LF situation
• handling scenario: if there are speed limits on the map, record it with continuous speed
• university track scenario: lane follow, with different speeds
• university track scenario: stop and go, traffic jam simulation

additional test cases:

• university track scenario (automated shuttle): cross-section simulation, stop point and cornering
• university track scenario (automated shuttle): parking scenario(s)
• university track scenario (automated shuttle): static obstacle avoidance with low speed

Data to be recorded:

• lidar point cloud
• camera image
• GNSS data
• map data (from file, stored in real time)
• vehicle kinematics

Create two mcaps, one with all data included and one with reduced size, necessary for LF step 1 simulations only

[gfigneczi1]

Test experience:

• inaccurate maps --> e.g., motorway, idea: record lane positions with in-vehicle camera and generate lanelet map out of it (direct link with GTC) - prerequisite: working camera; temporary solution: use Bosch car for data source. [mcap - python script - python library for building map file...]
• clean up test modifications: -- control command to actuator abstraction incl. enable flag -- foxglove fixing (add missing signals) + adding maximum speed as user input (instead of hard coded value) -- merge other changes -- check novatel orientation (inaccuracy and pi/2 rotation)
• finalize test node - add ldm planner
• standardize control node - add base class
• tf not ok problem