SACA Robotics and Automation 2 — Questions and Answers
Question 1: What defines the 'workspace' of an industrial robot arm?
- The floor space required for the robot's safety fence
- The three-dimensional volume that the robot's tool center point (TCP) can reach (Correct answer)
- The number of programs stored in the robot controller memory
- The maximum payload the robot can carry at full extension
Correct answer: The three-dimensional volume that the robot's tool center point (TCP) can reach
The workspace (or work envelope) is the complete set of positions in 3D space that the robot's tool center point can reach, determined by the robot's kinematic configuration, link lengths, and joint angle limits.
Different robot configurations produce different workspace shapes: articulated robots have a donut-shaped workspace (they cannot reach directly above or below themselves), SCARA robots have a cylindrical workspace, Cartesian robots have a rectangular prismatic workspace. The reachable workspace is the total volume the TCP can reach; the dexterous workspace is where the TCP can be positioned with any orientation. Workspace analysis is critical when integrating a robot into a cell — interference with fixtures, conveyor, and guarding must be verified.
Question 2: What is 'teach pendant programming' in industrial robotics?
- A simulation method using CAD models to generate robot paths offline
- Manually guiding the robot through desired positions while recording the joint angles for playback (Correct answer)
- Writing robot programs in a high-level language on a desktop computer
- Using a vision system to automatically teach robot positions from reference images
Correct answer: Manually guiding the robot through desired positions while recording the joint angles for playback
Teach pendant programming involves a technician physically jogging the robot arm to each required position using a handheld pendant controller, then recording those positions as waypoints that the robot replays during production.
The teach pendant is a handheld device with joysticks, axis keys, and a display. The technician jogs each axis individually or in Cartesian mode, positions the TCP at the desired point, then presses 'Record' to save the joint angles (point-to-point) or Cartesian coordinates. The robot program sequences through these taught points with motion commands (MOVJ for joint interpolation, MOVL for linear). Teach pendant programming is fast for simple pick-and-place but limited for complex curved paths, which require offline programming (OLP) with CAD simulation.
Question 3: What distinguishes a 'collaborative robot' (cobot) from a traditional industrial robot in terms of safety design?
- Cobots are faster and carry heavier payloads than traditional robots
- Cobots are designed with force/torque sensing and power limiting to safely stop when contact with a person is detected (Correct answer)
- Cobots use a different programming language exclusively developed for human interaction
- Cobots require no programming — they learn entirely from human demonstration
Correct answer: Cobots are designed with force/torque sensing and power limiting to safely stop when contact with a person is detected
Cobots (ISO/TS 15066) implement safety-rated force and power limiting, so they automatically stop or reduce speed when they detect unexpected contact — allowing them to work alongside humans without physical safety barriers in many applications.
ISO/TS 15066 defines four cobot operation modes: Safety-Rated Monitored Stop, Hand Guiding, Speed and Separation Monitoring, and Power and Force Limiting (PFL). PFL is the most common: joint torque sensors detect contact forces exceeding threshold (e.g., 150N) and halt within milliseconds. UR robots, FANUC CR series, and Kuka LBR iiwa implement PFL. A risk assessment (ISO 10218-2) is still required — even cobots can cause injury at high speed or with sharp tooling. The reduced safety barrier costs are offset by thorough risk assessment documentation.
Question 4: In industrial robot programming, what does 'tool center point' (TCP) calibration accomplish?
- It sets the maximum speed limit for the robot's end effector
- It defines the position and orientation of the active tool tip relative to the robot's mounting flange (Correct answer)
- It calibrates the robot's base coordinate system to the factory floor
- It programs the robot to automatically avoid fixture clamps in its path
Correct answer: It defines the position and orientation of the active tool tip relative to the robot's mounting flange
TCP calibration tells the robot controller exactly where the tool tip is relative to the robot's tool flange (mechanical interface), enabling the robot to control the tool tip position precisely and correctly during Cartesian moves and coordinated motions.
All Cartesian moves (MOVL, circular arcs) control the TCP position, not the flange. If the TCP is wrong (e.g., after changing a worn gripper), the robot positions its flange correctly but the actual tool tip is offset — causing parts to be picked or welded at wrong positions. TCP calibration methods: 4-point method (bring TCP to same Cartesian point from 4 different joint configurations, the controller computes the TCP offset), 6-point method (also calibrates orientation). After calibration, the TCP offset is stored in the tool data frame.
Question 5: What is 'dead reckoning' in autonomous mobile robot navigation, and what is its primary limitation?
- Using LIDAR to detect and map obstacles in real time; limited by sensor range
- Estimating current position by integrating wheel encoder data from a known starting point; limited by cumulative error accumulation (Correct answer)
- Using GPS coordinates for outdoor navigation; limited by indoor GPS unavailability
- Using pre-programmed waypoints without sensing; limited by dynamic obstacle handling
Correct answer: Estimating current position by integrating wheel encoder data from a known starting point; limited by cumulative error accumulation
Dead reckoning integrates wheel odometry (encoder counts × wheel circumference) from a known starting position to estimate current location. Small errors in each measurement accumulate over distance, causing growing position uncertainty.
Dead reckoning is the simplest AMR localization method: position = start + integral of velocity × time. Errors accumulate from: wheel slip (especially on turns), encoder resolution, unequal wheel diameters, and surface irregularities. A 1% wheel diameter error causes ~1% lateral drift — after 100m travel, position error can be >1m. AMRs combine dead reckoning with corrective sensors: LIDAR-based SLAM (simultaneous localization and mapping), magnetic floor tape, QR codes, or UWB beacons periodically correct the accumulated dead-reckoning drift.
Question 6: In industrial vision-guided robotics, what is the role of a 'calibration target' (e.g., checkerboard pattern)?
- To set the white balance of the camera for consistent color detection
- To establish the mathematical relationship between camera pixel coordinates and real-world robot coordinates (Correct answer)
- To calibrate the robot's force sensor after tool changes
- To verify that the vision light source intensity is within specification
Correct answer: To establish the mathematical relationship between camera pixel coordinates and real-world robot coordinates
A calibration target with known geometry (checkerboard, dot grid) allows the vision system to compute the camera-to-robot transformation — so when the camera detects a part at pixel (u,v), the system can compute the corresponding robot TCP position to pick it.
Camera calibration has two steps: 1) Intrinsic calibration — using many images of a checkerboard at different orientations to compute focal length, principal point, and lens distortion coefficients (Zhang's method, used in OpenCV and all major vision systems). 2) Hand-eye calibration — establishing the rigid transform between camera frame and robot base frame by recording robot poses and corresponding checkerboard detections. After calibration, when a vision system detects a part centroid at pixel coordinates, it applies the transform to compute the robot's required pick pose in base frame coordinates.
SACA Robotics and Automation 2
What defines the 'workspace' of an industrial robot arm?