Course on Masterclass Robotica
Section 1: Foundations of robotics [55 minutes]
An assistive laboratory robot must locate, transport and present scientific equipment to a human researcher. Practical activities use paper-based models, equipment cards and a laboratory floor plan; neither programming experience nor physical robotics hardware is required.
-
Analyse robotic systems and embodiment: Annotate a schematic of the laboratory robot, identifying its sensing, actuation, control and power subsystems; compare its mobile base and manipulator with an industrial arm, explaining how embodiment constrains operational capabilities. [10 minutes]
-
Apply principles of mechanisms, motion and actuation: Construct a two-link cardboard manipulator with pivot joints, trace its reachable workspace and compare alternative configurations for reaching an equipment location. Distinguish forward from inverse kinematics by predicting the endpoint from a configuration, then finding a configuration for a specified endpoint; assess the implications of payload and mechanical compliance. [15 minutes]
-
Analyse sensing and environmental perception: Examine prepared camera images, encoder readings and force measurements associated with the equipment-collection task. Classify the measurements as proprioceptive or exteroceptive, identify their respective contributions and annotate an equipment-location map where observations are noisy, ambiguous or incomplete. [15 minutes]
-
Apply feedback principles to behavioural regulation: Complete a tabletop control exercise using a target position of 50 cm, an initial measured position of 30 cm and successive correction cards. Record reference states, measured states and corrective actions; compare open-loop and closed-loop responses when an unexpected displacement occurs, identifying overshoot, responsiveness and disturbance rejection. [15 minutes]
Section 2: Reasoning and metacognition [55 minutes]
Building on the robot’s established physical and perceptual capabilities, the equipment-delivery task now requires explicit planning, uncertainty monitoring and researcher oversight.
-
Create representations and task plans: Convert the annotated laboratory map into object, location and constraint cards, then arrange an equipment-delivery sequence with explicit preconditions and anticipated outcomes. Distinguish task planning from motion planning by separating action selection from the specification of collision-free movement. [15 minutes]
-
Apply uncertainty-sensitive decision rules: Evaluate three successive decision cards involving an obscured equipment label, an unrecorded storage location and an uncertain grasp outcome. Using supplied confidence estimates and instructional decision thresholds, justify whether the robot should proceed, acquire further evidence or defer to the researcher; distinguish these classroom thresholds from validated operational safety criteria. [15 minutes]
-
Analyse metacognitive monitoring and calibration: Record confidence judgements before checking the outcomes of the preceding decisions, then compare confidence with correctness. Construct a monitoring table linking the robot’s predicted outcomes, observed discrepancies and proposed adjustments, distinguishing object-level execution from meta-level regulation and functional self-monitoring from subjective experience. [15 minutes]
-
Apply principles of human–robot collaboration and oversight: Role-play the next request in the same delivery sequence: “Bring the small flask”, when two candidates satisfy the description. Produce a clarification exchange that communicates uncertainty and intended action, then allocate authority for confirmation, interruption and recovery; justify how these choices support calibrated trust and accountable intervention. [10 minutes]
Section 3: Final review [10 minutes]
-
Synthesise key concepts: Annotate the completed equipment-delivery sequence to connect embodiment, sensing, feedback, planning, uncertainty monitoring and human oversight, using evidence from the practical exercises. [4 minutes]
-
Evaluate an integrated response: Diagnose a final discrepancy in which the robot reaches the designated location but cannot verify the equipment’s identity; specify the next observation, decision and human intervention, explaining their respective purposes without reconstructing the entire task plan. [4 minutes]
-
Formulate future learning directions: Identify one demonstrated capability and one unresolved question, then select a subsequent learning focus in robot programming, kinematics, perception, control or computational metacognition. [2 minutes]