MATLAB and Simulink project ideas
21 MATLAB and Simulink projects that Skill-Lync learners have actually built: ADAS features made the model-based design way, PID control of a mass-spring-damper and a DC motor, Stateflow logic, Simscape models, electric and hybrid vehicle models, and MATLAB code for engineering data and CFD. Each one comes with real reports you can read before you choose, the toolboxes you need, and one way to extend it into a final-year project.
As of 16 Nov 2026: 11,161 of 1,41,067 public learner reports match one of these ideas · Counts are reports, not unique students
How to pick one
- New to MATLAB? Start with a script. NASA data parsing and the Rankine cycle simulator need only base MATLAB (plus the free XSteam file) and teach file handling, loops and plots on numbers you can check against a steam table.
- Then model and control something. The mass-spring-damper and DC motor projects take you from equations to a transfer function to a tuned PID controller. This is the core of every controls interview.
- Aiming at automotive software jobs? Do one ADAS feature properly. Adaptive cruise control or traffic jam assist built from a requirement document, with a data dictionary, Model Advisor checks and generated code, is what model-based design (MBD) teams at OEMs and suppliers do every day.
ADAS and model-based design project ideas
Driver-assistance features built from requirement documents, the way automotive software teams build them.
Vehicle direction detection (MBD mini project)
The steering input arrives as a yaw rate and is compared with three calibration limits (left turn, right turn, straight) to decide which way the car is turning, with each requirement in its own subsystem and every signal named as the requirement document says. Small, but it is exactly how automotive software teams work.
Make it a final-year project: Add a requirement of your own (a lane-change flag when the turn lasts under 2 s), trace it in Requirements Toolbox and write the test that proves it.
Learn it in Introduction to Model-Based Development using MATLAB and Simulink
Adaptive cruise control in Simulink
Adaptive cruise control holds the driver's set speed and slows down to keep a safe gap when a slower car appears ahead. You build the feature from a requirement document the way OEM teams do: requirement tagging and traceability, a Simulink data dictionary (SLDD), Model Advisor checks, a 0.01 s sample time and C code from Embedded Coder.
Make it a final-year project: Add a stop-and-go mode below 30 km/h and test both modes on a recorded lead-vehicle speed trace.
Learn it in Introduction to Model-Based Development using MATLAB and Simulink
Traffic jam assist feature
Traffic jam assist drives off, follows, brakes and steers in slow, dense traffic while the driver supervises. Model the feature from its signal list, map the ports to AUTOSAR, run the modelling-guideline checks and generate the code.
Make it a final-year project: Write the hand-over logic: when and how the feature gives control back to the driver, and test every exit path.
Learn it in Introduction to Automotive Embedded Systems and AUTOSAR
Highway assist: lane changing assistant
On a highway the assistant handles speed and steering in monotonous driving. The lane change part decides, as a Stateflow state machine, when a change is allowed given the indicator, the blind spot and the cars around you, then the model goes through data dictionary, guideline checks and AUTOSAR code generation.
Make it a final-year project: Add a blind-spot sensor fault state and show the feature refuses a lane change when the sensor is unhealthy.
Learn it in Introduction to Automotive Embedded Systems and AUTOSAR
Rain-sensing wiper control and ABS vehicle speed
Two small ECU functions with a full verification workflow. The wiper runs at 100, 200 or 500 rpm for low, medium or high rain density from the rain sensor; the ABS part works out vehicle speed from the wheel speed sensors. Both go through a data dictionary, software-in-the-loop tests and guideline (JMAAB) checks.
Make it a final-year project: Add hysteresis so the wiper does not flick between speeds when the rain density sits on a threshold, and prove it with a test case.
Learn it in Software Verification and Validation and System Testing for Model-Based Development
TFT cluster speedometer software component
The instrument cluster gets vehicle speed from other ECUs over the CAN bus and decides what the driver sees. You turn the display requirements (filtering, unit conversion, valid ranges) into a Simulink model and take it through verification and validation, ending with a fixed-point model ready for a small microcontroller.
Make it a final-year project: Compare the floating-point and fixed-point models on the same speed trace and report the largest display error.
Learn it in Software Verification and Validation and System Testing for Model-Based Development
Control systems and physical modelling project ideas
Transfer functions, PID tuning, state machines and Simscape models.
Position control of a mass-spring-damper with PID
Write the transfer function of a mass-spring-damper (m = 3.6 kg, k = 400 N/m, c = 100 N·s/m), then tune a PID controller so the mass follows a pulse reference. With these numbers the system is overdamped (damping ratio about 1.32), so the open loop is slow but never overshoots: the controller's job is to make it fast without making it ring.
Make it a final-year project: Tune the gains three ways (by hand, Ziegler-Nichols, PID Tuner) and compare rise time, overshoot and control effort.
Learn it in Matlab and Simulink Basics
DC motor speed control from first principles
Write the armature circuit (Kirchhoff's voltage law with back EMF) and the torque balance, turn them into a state-space and transfer function model, and design a PID controller that brings the motor to a target speed.
Make it a final-year project: Add a load-torque step and an armature voltage limit, then add anti-windup and show the difference.
Learn it in Matlab and Simulink Basics
Washing machine and gear shift logic in Stateflow
Two state machines from written requirements: a washing machine (soak 200 s, wash 100 s, rinse 20 s, dry 50 s, stop with an LED if there is no water) and a gear selector (gear 1 up to 15 km/h, then 16-25, 26-40, 41-60 and above 61). Read the gear table closely: speeds between 15 and 16 km/h, or 60 and 61, are not covered, and a real speed signal passes through them.
Make it a final-year project: Fix the gaps, add a different up-shift and down-shift speed so the gearbox does not hunt, and write test cases that cover every transition.
Learn it in Introduction to Physical Modeling using Simscape
Doorbell solenoid and thermistor fan in Simscape
Physical modelling instead of equations: wire a doorbell from a switch, a solenoid and a spring, close the switch for 2 s and watch the plunger move. Then use a thermistor to switch a heater fan as the temperature steps between 20, 27 and 23 °C.
Make it a final-year project: Measure how plunger travel and response time change with supply voltage and spring stiffness.
Learn it in Introduction to Physical Modeling using Simscape
Vehicle and EV model project ideas in Simulink
Whole-vehicle models: electric cars, an airport tug, a hybrid's fuel economy and an e-rickshaw.
Design of an electric vehicle in Simulink
Build an electric car from Powertrain Blockset parts: a lithium-ion battery, a motor (DC or induction), the driveline, tyres and the vehicle body. Run it on a drive cycle and report the system configuration, parameters and results: speed tracking, battery current and state of charge.
Make it a final-year project: Model a real Indian EV from its published specs and compare your range on MIDC with the claimed figure.
Learn it in Introduction to Hybrid Electric Vehicle using MATLAB and Simulink
Electric powertrain for an aircraft towing vehicle
Aircraft are not taxied to the runway on their own engines; a tug pushes them. Work out the force and power needed to tow a given aircraft (rolling resistance, acceleration, grade), model it in Simulink, then size an electric powertrain (motor type, rating, energy) and the converter duty cycle that runs the tug from standstill to top speed.
Make it a final-year project: Size the battery for a full shift of tows at a real Indian airport and compare it with a diesel tug's fuel use.
Learn it in Introduction to Hybrid Electric Vehicle using MATLAB and Simulink
Forward energy-based fuel model: conventional vs P1 hybrid
A forward model has a driver block that presses the pedals to follow a drive cycle, so the powertrain behaves like a real one. Build it for a conventional car on UDDS, HWFET, FTP and NEDC, then add a motor on the crankshaft (a P1 mild hybrid) and work out the fuel economy gain in km/l.
Make it a final-year project: Write two energy-management rules for the hybrid and compare fuel economy and battery state of charge at the end of the cycle.
Learn it in Development of Hybrid Drives using MATLAB & Simulink
Electric rickshaw model in Simulink
Model a three-wheel passenger e-rickshaw with a rear PM brushed DC motor: vehicle, motor, controller and battery. Report energy use on three drive cycles and the temperature rise of the motor and controller at 45 km/h.
Make it a final-year project: Swap the brushed motor for a BLDC motor and compare efficiency and range, a practical question for Indian e-rickshaw makers.
Learn it in Introduction to Control of Electric Vehicle
MATLAB coding project ideas for engineers
File parsing, steam tables, optimisation and data handling.
Parsing NASA thermodynamic data
Read the NASA THERMO.dat file line by line, pull out the 14 polynomial coefficients for every species, and compute specific heat, enthalpy and entropy over each species' temperature range, plus its molecular weight. File parsing is the skill every CAE automation script needs.
Make it a final-year project: Turn it into a function that returns cp, h and s for any species and temperature, and check it against a published table for CO2 and H2O.
Learn it in MATLAB for Mechanical Engineers
Rankine cycle simulator with XSteam
Take the turbine inlet pressure and temperature and the condenser pressure (the course case is 30 bar, 400 °C, 0.05 bar), find all four state points with the free XSteam steam tables, and output net work, back work ratio, efficiency and the T-s and h-s diagrams.
Make it a final-year project: Add reheat or a feedwater heater and plot efficiency against boiler pressure.
Learn it in MATLAB for Mechanical Engineers
Optimisation with a genetic algorithm
The stalagmite function has many peaks, so a simple search gets stuck on a local one. Find its global maximum with MATLAB's ga, run it many times, and study how population size, generations and a fixed random seed change the answer.
Make it a final-year project: Use the same method on a real design problem, such as minimum-weight sizing of a beam with a stress limit.
Learn it in MATLAB for Mechanical Engineers
Data handling, curve fits and interpolation
Read measured voltage and current data from a spreadsheet with readtable, check it is numeric, fit a cubic with polyfit, interpolate at fixed steps with polyval, and write the result to a text file. The same course project also solves a spring-mass ODE with ode45 and ode23.
Make it a final-year project: Compare polynomial fitting with spline interpolation on the same data and show where each one goes wrong.
Learn it in Matlab and Simulink Basics
Numerical methods and CFD codes in MATLAB
Write the solver yourself: heat conduction, supersonic nozzle flow and finite-difference schemes.
Steady and unsteady 2D heat conduction solver
Write your own solver for heat conduction in a square plate with fixed edge temperatures: Jacobi, Gauss-Seidel and successive over-relaxation (SOR) for the steady case, explicit and implicit schemes for the transient one, and compare iterations and stability.
Make it a final-year project: Find the time step at which the explicit scheme blows up and check it against the stability limit from theory.
Learn it in Introduction to CFD using MATLAB and OpenFOAM
Quasi-1D supersonic nozzle flow with MacCormack's method
Simulate isentropic flow through a convergent-divergent nozzle that goes from subsonic to supersonic, with MacCormack's predictor-corrector scheme. Solve both the conservation and non-conservation forms of the equations and compare them with the exact solution.
Make it a final-year project: Do a grid independence study and plot the mass flow error along the nozzle for both forms.
Learn it in Introduction to CFD using MATLAB and OpenFOAM
Finite-difference schemes from a Taylor table
Derive 4th-order approximations of a second derivative: central (5 points, coefficients -1/12, 4/3, -5/2, 4/3, -1/12 divided by h²), skewed right and skewed left (6 points each), by solving the Taylor table as a linear system, then check the error falls with step size as the order predicts.
Make it a final-year project: Plot error against step size on log axes for all three and explain why round-off makes the error rise again at very small h.
Learn it in Introduction to CFD using MATLAB and OpenFOAM
Before you start
These are skill-building projects from Skill-Lync courses, not ready-made final-year submissions. Most start from course-provided data (the requirement documents, the THERMO.dat file, the mass-spring-damper values), so for a final-year project you need your own data or question, which is what the extension ideas are for.
The reports are learner submissions. Quality varies, some contain mistakes, and many ADAS reports show the model only as screenshots, so treat them as a view of the work, not as answers. Copying one will not get past a viva.
Counts match report titles, and some challenge titles are shared with other courses. We checked the tool in the 8 to 10 newest reports for every idea; any title that mentions Python, Cantera, C++ or Fortran is left out, and the pendulum ODE project is not listed because 4 of the 10 newest MATLAB-tagged reports were written in Python.
Questions students ask
Which MATLAB and Simulink project is best for a final-year student?
For automotive software and controls jobs, adaptive cruise control in Simulink built the MBD way: requirements, a Simulink data dictionary, Model Advisor checks and code generation. For mechanical roles, the forward energy-based fuel model of a conventional car and a P1 hybrid. Either one, extended with your own requirement or vehicle data, gives you something real to defend in a viva.
Where can I download an electric vehicle Simulink model?
MathWorks ships a full electric vehicle reference application with Powertrain Blockset, a paid add-on: motor, battery, direct-drive transmission, and a supervisory controller with the VCU and BMS. It is a good reference, but downloading a model teaches little. Building the 'design of an electric vehicle' project yourself, then comparing it with the reference, is what makes the work yours.
What is the difference between MATLAB and Simulink?
MATLAB is a programming language: you write scripts and functions, as in the NASA parsing, Rankine and genetic algorithm projects. Simulink runs on top of it and models systems that change over time as connected blocks, as in the motor, vehicle and ADAS projects. Automotive teams use Simulink because it can generate production C code from the model.
What is model-based design (MBD)?
A way of building embedded software where the Simulink model is the specification: each requirement is linked to the part of the model that meets it, the model is checked against modelling guidelines (MAAB or JMAAB) and tested, and the C code for the ECU is generated from it instead of written by hand. The six ADAS projects on this page follow that workflow.
Can I do these projects without a MATLAB licence?
The coding and numerical projects run in GNU Octave, which is free and reads most MATLAB code (the CFD course itself accepts MATLAB or Octave); the genetic algorithm needs Octave's ga package or your own GA. The Simulink, Stateflow and Simscape projects need Simulink: Octave has no equivalent, and free block-diagram tools such as Scilab's Xcos are not what automotive teams use.