Closed-loop motor control
Deliverable: A motor holding a commanded position under an applied disturbance, using encoder feedback, with the PID tuning process and final gains documented.
Course
Robotics
Robots built and tuned on a bench — drives, sensing, closed-loop control and ROS — because a machine that moves fails in ways no simulation shows you.
- Duration: 4 months
- Classroom
- Level: intermediate
What you will be able to do
- Choose motors, drivers and power for a stated load and speed
- Read encoders and implement closed-loop position and speed control
- Tune a PID controller on real hardware rather than in theory
- Fuse sensor readings to estimate position more reliably than either alone
- Build and run a basic ROS system with nodes and topics
- Test a robot safely, including what happens when something fails
Who this course is for
- Engineering students and graduates, mechanical, electrical or computer
- Automation technicians moving into robotics
- Anyone who has completed Arduino or an equivalent course
- Basic programming in C or Python is expected
Syllabus
7 modules · 4 months
Module 1. Mechanisms and drives
- Torque, speed and gearing
- Motor types — DC, stepper, servo, BLDC
- Selecting a motor and driver for a load
- Power budgeting and battery selection
Module 2. Motor control
- PWM and direction control
- Encoders and measuring actual movement
- Open loop versus closed loop
- Current limiting and stall protection
Module 3. Feedback and control
- Proportional, integral and derivative terms explained physically
- Tuning PID on hardware, methodically
- Overshoot, oscillation and steady-state error
- Why a controller tuned on a bench misbehaves on the floor
Module 4. Sensing and perception
- Ultrasonic, infrared and time-of-flight ranging
- IMU — accelerometer, gyroscope and drift
- Complementary and simple Kalman filtering
- Sensor limitations and where each one lies to you
Module 5. Navigation basics
- Odometry and its accumulating error
- Line following and wall following
- Obstacle avoidance and simple path planning
- Localisation, conceptually
Module 6. ROS fundamentals
- Nodes, topics, messages and services
- Publishing sensor data and subscribing to commands
- Visualising with RViz
- Simulation, and where it stops matching reality
Module 7. Safety and testing
- Emergency stop and safe failure
- Testing a moving machine without damage or injury
- Repeatability testing and logging
- Documenting a build
Tools and technologies you will use
- Arduino and Raspberry Pi
- DC motors, encoders and drivers
- Ultrasonic, IR and IMU sensors
- ROS on Ubuntu
- Multimeter and oscilloscope
Projects you will build
Sensor-guided robot
Deliverable: A mobile robot that follows a line or wall reliably over ten consecutive runs, with failures logged and the cause of each identified.
Sensor fusion exercise
Deliverable: Heading estimated from a gyroscope alone, an accelerometer alone, and both fused, with the drift of each compared over five minutes of recorded data.
ROS system
Deliverable: A working ROS setup with at least three nodes publishing and subscribing, visualised in RViz, running on the physical robot rather than in simulation only.
Where this course can take you
- Robotics Engineer
- Automation Engineer
- Mechatronics Engineer
- Product Development Engineer
- Robotics Trainer
Duration, modes and fees
- Duration
- 4 months
- Delivery modes
- Classroom
- Fees
- Share your details for the current fee
- Fees vary by batch. Robot platforms, motors, sensors and lab equipment are provided. Classroom mode only.
Placement assistance
Every student gets placement assistance — that is what 100% placement assistance means. It is support for all, not a job for all. We do not promise a specific salary, a specific number of interviews, or placement at any named company, and you should be wary of anyone who does.
What is included
- A place in the monthly placement drive, held every third Saturday
- The readiness programme every second Saturday — mock interviews and preparation
- CV review against the specific roles you are targeting
- Portfolio review, so your project work is presented the way a reviewer will read it
- Access to the vacancy pool employers send directly to the Samyak network
- Guidance on which roles realistically fit your background and which do not
- A place in the next drive, with coaching, if you are not selected in this one
What is not included
- Any guarantee of a job, an interview, or a particular salary
- Placement at a named or partner company
- Applying to jobs on your behalf
- Support before you have completed the course and its project work
- Visa, relocation or overseas placement assistance
Things that move fail differently
Software fails and you read a log. A robot fails and something hits a wall, a gear strips, or a motor draws stall current until the driver gets hot.
That changes how you work. You test incrementally, you power through a current limit, you keep a hand on the stop, and you assume every new piece of code will do the wrong thing the first time. The safety module is not paperwork — it is how the lab actually runs.
Tune on the bench, not in the model
A PID tuned in simulation will very often oscillate on real hardware.
Friction, backlash, flex in a printed bracket, a battery sagging under load — the model has none of it. So tuning is taught as a method carried out on a moving machine: change one term, watch what happens, write it down. Students find the derivative term amplifying encoder noise on their own, which is worth more than being told it might.
Every sensor lies somewhere
Ultrasonic ranging fails on soft surfaces and at an angle. Infrared is fooled by sunlight. A gyroscope drifts steadily. An accelerometer is noisy the instant anything vibrates.
The fusion project makes this concrete: estimate heading three ways and watch the gyro drift away over five minutes while the fused estimate holds. Knowing where each sensor is unreliable is most of what perception actually is at this level.
ROS after the fundamentals, not before
You can build a working robot without ROS and the first four modules do.
Once there are several sensors, several processes and something you would like to reuse, ROS starts earning its complexity. Meeting it in that order means it is organising things you already understand rather than being a framework you believe in.
Questions
Robotics — frequently asked questions
Do I need mechanical or electronics background?
Either helps and neither is sufficient alone — robotics sits across mechanical, electrical and software, and most people arrive strong in one. The course builds the other two enough to be dangerous in the right way. Completing Arduino or an equivalent first makes it considerably smoother.
Why tune PID on hardware rather than in simulation?
Because real systems have friction, backlash, flex and a power supply that sags under load, and none of that is in your model. A controller tuned in simulation routinely oscillates on the bench. Tuning by method on real hardware, watching what each term does, is the skill this course is built around.
Is ROS necessary for a beginner?
Not to make a robot move, and the first four modules do not use it. It becomes necessary the moment you have several sensors, several processes and want to reuse anything. Introducing it after the fundamentals means you understand what it is organising rather than treating it as a magic framework.
How much does the hardware cost?
Lab hardware is provided and included in the fee, which is why this is classroom-only. If you want to continue at home afterwards, a capable starter platform is affordable, and we will tell you honestly what is worth buying and what is not.
Enquire about Robotics
Three details is all we need. A course advisor will call you back.
Related courses
- Arduino
The fastest way to make electronics actually do something — sensors, motors, displays and code — taught by building prototypes you design rather than diagrams you copy.
- PLC and SCADA Automation
Industrial automation taught on real hardware — ladder logic, analogue I/O, HMI and SCADA — with the interlock and safety discipline that separates a working panel from a dangerous one.
- Raspberry Pi
A small Linux computer used properly — Python, GPIO, camera and networking — building headless devices that start on their own and keep running unattended.
- IoT (Internet of Things)
Connected hardware from sensor to dashboard — microcontrollers, wireless protocols, cloud ingestion and the power and reliability problems that decide whether a device survives deployment.
Next step
Talk to a course advisor
Tell us what you want to learn and we will help you pick the right course, batch and mode.