AI & Robotics Foundations

When does a programmed machine become an AI-enabled robot?

• Distinguish automation, robotics, AI, and embodied AI.
• Trace a complete sense-estimate-plan-act-check loop.
• Identify autonomy boundaries and human oversight points.

How does a robot connect software decisions to the physical world?

• Explain the responsibilities of sensors, actuators, compute, power, and mechanics.
• Compare sensors using range, resolution, latency, and failure modes.
• Select components for a constrained robot task.

How does a robot describe where it is and where it should move?

• Represent position and orientation in named coordinate frames.
• Relate velocity, acceleration, degrees of freedom, and constraints.
• Apply simple forward and inverse kinematics.

How do independent robot components communicate without becoming one giant program?

• Explain nodes, messages, topics, services, and actions.
• Select the appropriate ROS 2 communication pattern for a task.
• Inspect and diagnose a small robot computation graph.

Why does a robot need to keep checking its actions?

• Distinguish open-loop and closed-loop control.
• Explain setpoint, error, gain, overshoot, and stability.
• Tune and evaluate a simple feedback controller.

How does a robot turn sensor measurements into useful evidence?

• Explain images, depth, filtering, and visual features as robot data.
• Compare classification, detection, segmentation, pose, and tracking.
• Evaluate a perception pipeline using errors and changing conditions.

How can a mobile robot move when its map and location are uncertain?

• Explain odometry, drift, landmarks, and state estimation.
• Interpret an occupancy map and the purpose of SLAM.
• Diagnose failures across localization, planning, and control.

How does a robot choose what to do next?

• Compare reactive rules, state machines, and behavior trees.
• Explain graph search, costs, constraints, and collision avoidance.
• Design fallback and replanning behavior under uncertainty.

What changes when a robot must touch objects or work near people?

• Explain grasping, reachability, collision checking, and contact.
• Sequence perception and motion for a pick-and-place task.
• Design clear intent signals and human override points.

What should a robot learn, and what should remain explicitly engineered?

• Explain supervised, imitation, and reinforcement-learning roles in robotics.
• Separate training performance from deployment evaluation.
• Identify reward, data, distribution-shift, and sim-to-real risks.

Where can generative AI help a robot, and where should it not have direct authority?

• Explain grounding, affordances, and task decomposition.
• Design a tool-using robot agent with verification boundaries.
• Identify hallucination, prompt-injection, and permission risks.

How do we demonstrate that a robot is useful, reliable, and appropriately governed?

• Apply safety, security, privacy, fairness, and accountability criteria.
• Build a reproducible test matrix with meaningful metrics.
• Present an integrated robot mission with limitations and human oversight.