Unit 8: Programming the Robot - Subjective Questions

Robot programming is the process of creating a sequence of instructions that enables the Sierena's Utility Robot U-BOT to sense its surroundings, make decisions, and perform physical actions.

Its importance includes:

• Task automation: The robot can perform color identification, shape identification, and path-tracking tasks automatically.
• Sensor integration: Programs process data obtained from color, optical, infrared, or other sensors.
• Motion control: Instructions determine the speed, direction, and duration of motor movement.
• Decision-making: Conditional statements allow the robot to respond differently to different colors, shapes, or path conditions.
• Repeatability: A programmed robot performs the same task consistently and accurately.
• Safety and efficiency: Proper programming reduces unnecessary movement, collisions, and human intervention.

The basic procedure consists of the following steps:

• Understand the task: Identify the required input, robot action, and expected output.
• Select sensors and actuators: Determine which sensors will detect colors, shapes, or paths and which motors will move the robot.
• Develop the algorithm: Write the sequence of operations using a flowchart, pseudocode, or programming language.
• Initialize the robot: Configure communication, sensor ports, motor ports, speed values, and threshold values.
• Implement the program: Convert the algorithm into instructions supported by the U-BOT programming environment.
• Compile or verify: Check the program for syntax and logical errors.
• Upload the program: Transfer the code to the robot controller.
• Test in a controlled area: Observe sensor readings and robot movement.
• Calibrate and modify: Adjust thresholds, speed, and timing based on the test results.
• Run the final task: Execute the program and record the performance.

Color identification is programmed by using a color sensor or a sensor arrangement that measures the reflected or received light from an object.

The general process is:

• The sensor is positioned at a suitable distance from the object.
• The robot reads the reflected intensity or color values, such as red, green, and blue components.
• The program compares the sensor readings with calibrated reference values.
• Conditional statements classify the object as a particular color.
• The robot performs a predefined action, such as stopping, turning, sorting, or displaying the detected color.

A simplified decision structure is:

• If the reading matches the red range, classify the object as red.
• Else if the reading matches the green range, classify it as green.
• Else if the reading matches the blue range, classify it as blue.
• Otherwise, classify it as unknown.

Calibration is essential because ambient light, surface texture, sensor distance, and object orientation can change the readings.

Calibration establishes reliable sensor values for each color under the actual laboratory conditions.

The procedure is:

• Place the robot and color sensor in the same position that will be used during operation.
• Keep the sensor at a fixed distance from the test surface.
• Measure the sensor response for a reference background or neutral surface.
• Place standard samples of each required color under the sensor.
• Record multiple readings for every color to account for small variations.
• Determine a suitable range or threshold for each color.
• Repeat the measurements under the expected lighting conditions.
• Store the calibrated values in the program.
• Test the program with additional samples that were not used during calibration.
• Modify the thresholds if the robot confuses two colors.

Calibration should be repeated whenever the sensor position, lighting, surface, or operating environment changes.

A suitable algorithm is:

1. Start the robot.
2. Initialize the color sensor and motors.
3. Move toward the object at a controlled speed.
4. Read the color sensor values.
5. Compare the values with calibrated color ranges.
6. If the object is red, stop and move to the red collection area.
7. Else if the object is green, stop and move to the green collection area.
8. Else if the object is blue, stop and move to the blue collection area.
9. Else, classify the object as unknown and move to a reject area.
10. Return to the starting position or wait for the next object.
11. Stop.

The algorithm should include a tolerance range and a short confirmation delay so that a single noisy sensor reading does not produce an incorrect classification.

Shape identification is the process of determining the geometric form of an object, such as a circle, square, rectangle, or triangle, using sensor data.

The U-BOT may identify shapes using:

• Distance sensors: Multiple measurements can reveal the distance between the robot and different parts of an object.
• Color or optical sensors: A sensor can scan the boundary of a shape and detect changes between the object and background.
• Touch sensors: Contact points can be recorded while the robot moves around an object.
• Encoders: Wheel movement can estimate the distance traveled along each side.
• Camera-based sensing: Images can be processed to detect edges, corners, and contours if a vision system is available.

For example, a square generally produces four corners and four approximately equal sides, while a circle produces a continuously changing boundary without sharp corners. The program compares measured features with predefined shape conditions and assigns the most appropriate shape label.

Color identification and shape identification differ mainly in the type of information being measured.

• Color identification determines the spectral or reflected-light properties of an object. It commonly uses a color sensor and compares measured values with calibrated color ranges.
• Shape identification determines the geometric properties of an object. It may use distance, optical, touch, encoder, or camera data to identify edges, corners, sides, or contours.

The main differences are:

• Color identification depends strongly on lighting, surface reflectivity, and sensor distance.
• Shape identification depends on object boundaries, dimensions, orientation, and the resolution of measurements.
• Color classification usually requires comparison with color thresholds.
• Shape classification usually requires analysis of geometric features.

Both methods use sensing, data processing, decision logic, and an output action. They can also be combined, such as identifying a red circle or a blue square.

A shape-identification program can be developed by defining measurable features for each shape.

• Square: Detect four corners and four sides of approximately equal length.
• Circle: Detect a smooth boundary with no sharp corners and approximately constant distance from the center.
• Triangle: Detect three corners and three sides.

A general method is:

1. Move the sensor around or across the object in controlled steps.
2. Record distance or boundary readings at each step.
3. Filter abnormal readings caused by noise.
4. Detect sudden changes that indicate corners or edges.
5. Count the number of corners.
6. Compare side lengths and boundary characteristics.
7. Assign the shape whose features best match the measured data.
8. Report or act on the identified shape.

The program should use tolerances because real sensor measurements are not perfectly exact. The object must also be placed within the sensor's operating range.

A threshold is a reference value used to convert a continuous sensor measurement into a decision or category.

In color identification, thresholds define the acceptable range for a color. For example, readings within a calibrated range may be classified as red, while readings outside that range may be classified as another color or as unknown.

In shape identification, thresholds can define:

• The minimum change required to detect an edge.
• The distance variation considered to be a corner.
• The allowable difference between side lengths.
• The tolerance for deciding whether a boundary is circular or straight.

Thresholds must be selected carefully:

• A threshold that is too low may respond to sensor noise.
• A threshold that is too high may miss real changes.
• Fixed thresholds may fail when lighting or object position changes.
• Calibrated ranges and hysteresis can improve reliability.

Thus, thresholds connect raw sensor readings to meaningful robot decisions.

Path tracking is the ability of a robot to follow a predefined route or guide, such as a line, colored strip, tape path, or marked track.

It is important because:

• It allows the U-BOT to move between known locations automatically.
• It supports material handling and transport applications.
• It reduces the need for manual control.
• It improves repeatability of movement.
• It helps the robot avoid leaving the intended route.
• It provides a practical application of sensor feedback and motor control.

A path-tracking system continuously senses the position of the path relative to the robot. The controller then adjusts the left and right motor speeds to reduce the tracking error. The process repeats until the robot reaches the end of the path or receives a stop condition.

A line-following program uses optical or infrared sensors to distinguish a path from its background.

The working principle is:

• Sensors are mounted near the front or underside of the robot.
• The sensors detect reflected light from the surface.
• A dark line and a light background produce different sensor readings.
• The program determines whether the line is to the left, centered, or right of the robot.
• Motor speeds are adjusted to steer the robot back toward the center of the line.

A typical control response is:

• Line on the left: Increase the right motor speed or reduce the left motor speed.
• Line centered: Drive both motors at approximately equal speed.
• Line on the right: Increase the left motor speed or reduce the right motor speed.
• No line detected: Stop, search for the line, or execute a recovery routine.

The sensor threshold must be calibrated for the track material and lighting conditions.

Both methods use multiple sensors to determine the position of a line, but they provide different levels of information.

Two-sensor method:

• Uses a left sensor and a right sensor.
• Has a simple hardware and programming structure.
• Can generally detect whether the line is left, right, or between the sensors.
• May have difficulty distinguishing sharp turns, intersections, and complete line loss.

Three-sensor method:

• Uses left, center, and right sensors.
• Can identify a centered line more directly.
• Provides better detection of moderate deviations and some intersections.
• Requires additional input processing and calibration.

The three-sensor method usually gives smoother and more reliable path tracking because it provides more positional information. However, the best choice depends on the track design, available hardware, required speed, and program complexity.

Motor speed control changes the movement of the left and right wheels according to the sensor-detected position of the path.

• If the path shifts to the left, the robot changes the relative wheel speeds to turn left.
• If the path shifts to the right, the relative wheel speeds are changed to turn right.
• If the path is centered, both motors run at similar speeds.

The speed setting affects performance:

• Very high speed: The robot may overshoot turns, oscillate around the line, or lose the path.
• Very low speed: Tracking becomes slow, although it may be more stable.
• Unequal motor speeds: The robot may drift even when the sensors indicate a centered path.
• Rapid corrections: The robot may vibrate or produce unstable steering.
• Insufficient correction: The robot may gradually leave the path.

A suitable base speed and correction amount must be selected through testing. Motor calibration is also necessary because two motors may not produce identical motion for the same command.

Let the desired path position be the center position and let the measured path position be obtained from the sensors. The tracking error is defined as:

A proportional controller generates a correction based on this error:

where is the proportional gain and is the steering correction.

The motor commands can be written as:

Here, is the normal forward speed. When the error is zero, both motors receive approximately equal commands. When the robot moves away from the center, the difference between the motor speeds turns it back toward the path.

A small value of may cause slow correction and path drift. A large value may cause oscillation. The gain should therefore be adjusted experimentally, and motor commands should be limited to their safe operating range.

Sensor placement and mechanical alignment directly affect the quality of path-tracking decisions.

Important considerations include:

• Sensors should be mounted at a fixed height above the track.
• The sensing elements should be aligned parallel or perpendicular to the path as required by the design.
• The distance between sensors should be selected so that small deviations can be detected.
• Sensors should be positioned far enough ahead of the drive wheels to allow correction before the robot crosses the path.
• The sensor bracket must be rigid so that vibration does not change the readings.
• The robot chassis and wheels should be aligned to prevent continuous drifting.
• The sensing area should be protected from excessive ambient light when necessary.

Poor alignment can cause unequal readings, late corrections, false line detection, and unstable motion even when the program logic is correct.

Color identification and path tracking can be combined by assigning special meanings to colored sections or markers on a path.

A possible program sequence is:

1. Start line tracking using the optical or infrared path sensors.
2. Continuously read the color sensor while the robot moves.
3. If no special color is detected, continue following the path.
4. If a red marker is detected, stop and perform the red-marker action.
5. If a green marker is detected, turn toward a specified route or release an object.
6. If a blue marker is detected, record the event or move to a different destination.
7. Resume path tracking after completing the selected action.
8. Stop when the final marker or end-of-path condition is detected.

The program should prevent repeated detection of the same marker by using a delay, state change, or requirement that the robot move beyond the marker before checking again.

A state-based program divides the robot's behavior into clearly defined operating modes. At any time, the robot executes the instructions associated with its current state and changes state when a specified event occurs.

Suitable states include:

• Initialize: Configure sensors, motors, thresholds, and variables.
• Search: Look for the path or target object.
• Track: Follow the path using sensor feedback.
• Detect: Read sensor data and identify a color or shape.
• Classify: Compare measured features with stored conditions.
• Act: Stop, turn, sort, display, or otherwise respond to the result.
• Recover: Search again if the path or object is lost.
• Finish: Stop the robot after the task is complete.

State-based programming improves organization, debugging, and reliability. It also prevents unrelated actions from being executed simultaneously and makes transitions explicit.

Common sources of error include:

• Ambient light: Changes the sensor response and may cause incorrect color classification.
• Sensor distance: A different distance changes the reflected intensity.
• Surface finish: Glossy, rough, or transparent surfaces produce different readings.
• Object orientation: A rotated or tilted object may expose different features.
• Sensor noise: Random fluctuations may create false detections.
• Insufficient calibration: Reference values may not represent actual operating conditions.
• Overlapping thresholds: Similar colors or shapes may satisfy more than one condition.
• Motor vibration: Movement can alter the sensor position during measurement.

Error reduction methods include:

• Calibrating under actual lighting conditions.
• Taking multiple readings and using an average or majority decision.
• Applying suitable thresholds and tolerances.
• Filtering abnormal readings.
• Fixing the sensor mount and operating distance.
• Using an unknown category when confidence is low.
• Testing with different samples and orientations.

In open-loop control, the robot sends motor commands without measuring whether the desired movement has occurred. For example, it may command both motors to run for a fixed time and assume that the robot has traveled the required distance.

In closed-loop control, sensors measure the robot's actual condition or position. The controller compares the measured result with the desired result and adjusts the motor commands.

Closed-loop control is preferred for path tracking because:

• It responds to changes in the robot's position relative to the path.
• It corrects errors caused by wheel slip or unequal motor performance.
• It adapts to small changes in speed and surface conditions.
• It can recover from deviations and partial line loss.
• It provides better accuracy than fixed timing alone.

Path sensors provide feedback, while the motor controller uses that feedback to produce corrective steering.

A systematic debugging procedure is:

• Inspect hardware: Check sensor connections, motor wiring, battery level, wheel condition, and mechanical alignment.
• Test sensors separately: Display or record raw sensor readings for the path, background, and each color sample.
• Verify calibration: Confirm that thresholds match the current lighting, surface, and sensor distance.
• Test motor directions: Ensure that a forward command moves both wheels forward and that left and right corrections are correct.
• Use a low speed: Run the robot slowly to separate sensor errors from motion instability.
• Check decision logic: Test every color, shape, line position, and no-detection condition.
• Add temporary diagnostic output: Record sensor values, detected classifications, and current program state.
• Test one function at a time: Validate color detection, shape detection, and path tracking independently before combining them.
• Adjust parameters: Modify thresholds, sensor placement, base speed, and correction gain.
• Repeat under realistic conditions: Test with different objects, orientations, path sections, and lighting conditions.

The final program should be tested repeatedly after each significant modification.