AI Planner Algorithm Engineer_XC

Role Summary

L2 level longitudinal and lateral planning control algorithm development: Responsible for the planning and control strategy design of functions such as ACC (Adaptive Cruise), LKA (Lane Keeping Assist), TJA (Traffic Jam Assist), etc., focusing on optimizing driving comfort (such as jerk suppression, smooth cornering, soft starting/stopping when following a car).

Application of AI methods in comfort optimization: Explore imitation learning, reinforcement learning or lightweight end-to-end models to learn smooth throttle/brake/steering curves from human driving data to reduce the mechanical feel caused by rules.

Joint balance of comfort and safety: While ensuring priority in triggering of safety functions such as AEB, it optimizes false triggers (such as false deceleration, unnecessary alarms) and excessive intervention in daily driving to enhance user trust.

Integration of functional safety and comfort constraints: Design the redundancy and monitoring mechanism of the control module in accordance with ISO 26262 (ASIL level requirements) to ensure smooth degradation without sudden actions under abnormal conditions (such as sensor failure, signal frame loss).

Simulation and real vehicle closed-loop evaluation: Build an automated evaluation system that includes comfort indicators (impact, lateral acceleration change rate, vehicle following deviation), and use road mining data to drive iterations.

Requirements

Basic requirements

• Vehicle engineering, control engineering, computer, automation and other related majors, bachelor degree or above, more than 2 years of experience in L2 ADAS regulation and control development.
• Proficient in C++/Python and familiar with Linux and embedded platforms such as Infineon, NXP, TI TDA4 or Horizon Journey.
• Familiar with classic planning control algorithms: PID, MPC, LQR, pure tracking control, Lattice Planner or dynamic programming/quadratic programming.

Extra points/special requirements

• Has experience in mass production of L2 functions (ACC, LKA, TJA), and has taken the lead in comfort adjustment (such as corner speed control, start/stop smoothness).
• Understand the application of AI in regulation: such as using behavioral cloning to optimize car following strategies, or using reinforcement learning to learn lane-changing comfort parameters.
• Familiar with functional safety standards: Understand ASIL classification, FFI analysis, and the implementation of FMEA in the regulation and control module.
• Have good data-driven thinking: able to use large-scale road mining logs for problem location and regression testing.