Engineering Systems Roadmap
roadmap, engineering systems, control, signal processing, scientific computing
1 Purpose
This roadmap is for readers who want math to feel connected to systems that evolve, sense, estimate, communicate, and act.
It is not a pure-proof route and it is not a narrow controls curriculum.
It is a dependency-aware path through the parts of the site that support:
- dynamical systems
- sensing and signal pipelines
- feedback and optimal control
- stochastic decision-making
- computation and simulation
2 Who This Is For
Use this roadmap if your goal is any of the following:
- understand state-space models, stability, and feedback
- read signal, sensing, or communication papers without treating the math as black boxes
- connect numerical methods to simulation, inversion, and large systems
- move toward robotics, estimation, scientific computing, or sequential decision-making
If your goal is mostly statistical learning theory or proof-heavy ML theory, AI / ML Theory is the cleaner roadmap.
3 Main Sequence
Use this as the default order.
- Algebra Repair
- Linear Algebra
- Single-Variable Calculus
- Multivariable Calculus
- Probability
- Statistics
- ODEs and Dynamical Systems
- Numerical Methods
- Matrix Analysis
- Signal Processing and Estimation
- Control and Dynamics
- Stochastic Control and Dynamic Programming
- Information Theory
That sequence is now largely live on the site end to end, including complete first-pass modules for ODEs and Dynamical Systems, Numerical Methods, Matrix Analysis, Signal Processing and Estimation, Control and Dynamics, Stochastic Control and Dynamic Programming, and Information Theory.
4 Why This Order Works
4.1 Linear Algebra And Calculus First
Engineering systems keep reusing the same objects:
- vectors and matrices
- derivatives and local approximation
- gradients, Jacobians, and linearization
- quadratic forms and spectral structure
Without that language, state-space models, filters, and operators become notation to memorize instead of reusable tools.
4.2 Dynamics Before Feedback
Feedback only makes sense once a reader can describe how a system evolves on its own.
That is why ODEs and Dynamical Systems comes before Control and Dynamics.
First learn:
- state variables
- solution curves
- equilibria
- local stability
Then it becomes natural to ask how inputs and feedback change that behavior.
4.3 Numerical Methods Before Large-Scale Systems Work
Real systems work is rarely only symbolic.
It depends on:
- solving linear systems
- computing eigenstructure
- simulating trajectories
- handling least squares and regularization
- controlling numerical error
That is why Numerical Methods belongs in the core systems path instead of as an optional afterthought.
4.4 Probability And Statistics Before Estimation Under Uncertainty
Signals, sensors, and controllers rarely operate in perfect conditions.
Noise, hidden state, uncertainty, and limited data force a reader to think probabilistically.
That is why Probability and Statistics come before the deeper estimation and stochastic-control material.
6 Branch Points
After the shared core, most readers should branch by goal instead of forcing one long linear route.
6.1 Control And Robotics Branch
Use this branch if you care about:
- feedback
- stability
- controllability and observability
- trajectory planning
- MPC and constrained control
Best current pages:
6.2 Sensing And Communication Branch
Use this branch if you care about:
- filtering and denoising
- sampling and bandwidth
- decoding and error tradeoffs
- sensing and inverse reconstruction
- information limits
Best current pages:
6.3 Scientific Computing Branch
Use this branch if you care about:
- stable simulation
- numerical linear algebra
- time-stepping
- inverse problems
- approximation and error control
Best current pages:
6.4 Stochastic Decision-Making Branch
Use this branch if you care about:
- noisy dynamics
- Bellman methods
- partial observability
- LQG and filtering
- RL/control bridges
Best current pages:
7 Pages On The Site That Already Support This Roadmap
7.1 Strongest Current Systems Math Support
7.2 Strongest Current Systems Bridge Pages
8 Paper Reading Overlay
Do not wait until the whole roadmap is complete before touching systems papers.
Run a light reading overlay in parallel:
- How to Read a Paper
- one application-facing systems bridge page
- one matching advanced module page
A good current sequence is:
9 Common Next Directions
After the current route, the strongest adjacent additions would be:
Stochastic Processes, for Brownian motion, SDEs, MCMC, and diffusion-side probabilityApplications > Scientific Computing, for simulation and PDE-facing bridgesApplications > Optimization and Inference, for inverse problems, variational methods, and estimator-design workflows
10 Sources and Further Reading
- 18.03 Differential Equations syllabus -
First pass- official MIT anchor for the continuous-time modeling side of systems thinking. Checked2026-04-25. - 6.003 Signals and Systems lecture notes -
First pass- official MIT signal-and-system anchor for channels, convolution, and acquisition language. Checked2026-04-25. - 6.241J lecture notes -
First pass- official MIT state-space and control anchor. Checked2026-04-25. - 16.323 lecture notes -
Second pass- official MIT optimal-control anchor once planning and stochastic control begin to matter. Checked2026-04-25. - EE278 / Introduction to Statistical Signal Processing -
Second pass- official Stanford estimation anchor for noisy sensing and hidden-state inference. Checked2026-04-25. - EE363 bulletin -
Second pass- official Stanford control anchor for linear systems and feedback language. Checked2026-04-25. - AA228 / Decision Making Under Uncertainty -
Paper bridge- strong Stanford bridge once stochastic control, planning, and partial observability begin to overlap. Checked2026-04-25.