Who Owns What?
Ever wondered what people were up to? This page is kept up to date as projects grow and new features are added so people know whats going on!
Dynamics Controller
Section titled “Dynamics Controller”Low-level vehicle controller. Takes in driver/driverless input and commands torques to each wheel. Works closely with Firmware!
| Project | Description | Owners |
|---|---|---|
| General Vehicle Controller R&D | R&D into a general vehicle controller that can be run for all events, including driverless. This project is focused on our core torque vectoring algorithm and the installation of a driverless interface for low-level controls. | Arnav (lead), Chris, David, Joseph, Animesh, Zayan, Justin, Sid, John |
| Regenerative Braking R&D | Regen braking will be a large focus for 27x, since braking from the pedal will blend both regen and hydraulic capabilities. This project encompasses the R&D required to explore, integrate, and tune various regen strategies for next year’s competition vehicle. | Sam (lead), Lana, Hani, Siddhant, Harish |
| Monorepo QoL | Improve the safety, quality, and readability of the controls codebase within monorepo. | Sid, Ekaansh |
Driverless Controller
Section titled “Driverless Controller”High-level driverless trajectory planner. Currently, this refers to our in-house implementation of MPPI, a sampling-based MPC!
| Project | Description | Owners |
|---|---|---|
| Performance & Profiling | Effort to squeeze the most performance out of our driverless algorithms, the focus of P&P is to own the tradespace and ultimately make our code faster and safer! | Husain (lead), Ronav, Siddhant, Poon B |
| Algorithm R&D | R&D into improving the optimality, expressiveness, and stability of our trajectory planner via algorithm exploration and validation. Tickets will likely focus on sampling and cost function optimizations. | Richard (lead), Lucy, Ishaan, Eugene, Angela, Ekaansh, Sam |
| Driverless QoL | General codebase improvements with a focus on making our driverless codebase cleaner and easier to work with. | Animesh, Joaquin, Lucy |
CI Testing & Simulation
Section titled “CI Testing & Simulation”Testing harnesses and simulators for validating controls code. Works closely with DV Software Architecture!
| Project | Description | Owners |
|---|---|---|
| Carmaker HiL/SiL | Development of dedicated software-in-the-loop and hardware-in-the-loop testing harnesses for EV, DV, and unified controls testing within IPG Carmaker. | Richard, Joseph, Devon, Zayan, Hani |
| DV Controls Simulator | Bring-up and QoL improvements for our in-house driverless trajectory planning simulator. A large focus is integrating the simulator with CI/CD to allow automated acceptance testing of the driverless trajectory planner. | Harish, Joyce, Angela |
Dynamics Modeling
Section titled “Dynamics Modeling”Vehicle models, system identification, and state projection. Works closely with Vehicle Dynamics!
| Project | Description | Owners |
|---|---|---|
| System Identification & Tooling | Data analysis & infrastructure build-out to perform real-time dynamics analysis, post-run correspondences, and determine sim-to-real error boundaries. | Lana, David, Justin, Brandon |
| Vehicle Modeling R&D | Development, validation, and implementation of models for a variety of vehicle systems, from motor efficiency lookup tables, to suspension response under transient loads, and even autonomous steering propagation delay. The goal is not only to formulate models for dynamic systems, but to verify and solidify their implementations in code. | Devon, Brandon, Joyce, Joaquin, Poon B |
| State Projection QoL | State projection accounts for latency differences across the driverless pipeline and advances previously observed/computed states. This code is currently heavily intertwined with the core trajectory planning algorithm, making it hard to characterize and reason about. This project takes ownership of and decouples the state projector, allowing for higher-fidelity vehicle models and better observability. | Ishaan, Eugene, John |
