Priya's Wednesday, two ways

Priya, a robotics engineer at a mid-size manufacturer.

Priya starts Wednesday with a six-axis arm on the packaging line that keeps missing its pick position. Before the course, she would spend the first hour scrolling through ROS logs and guessing at the cause. Today she opens Claude and uses the Troubleshoot Robot Behavior prompt. She pastes a short description of the arm's motion, the gripper sensor readings, and a photo of the error screen. The AI gives her three ranked hypotheses and two quick checks. She finds a loose encoder cable in twenty minutes.

Next she needs to adjust the control loop for a new conveyor speed. She uses the Draft Control System Code prompt in ChatGPT. She writes: 'ROS node, Python, simple state machine for pick and place, conveyor speed 0.3 m/s, gripper open and close.' The AI returns starter code with comments. Priya reads it, fixes two lines, and runs it in simulation.

Before touching the real arm, she uses the Simulate Before You Build prompt in Gemini. She asks for a stress scenario where the conveyor jams and the arm has to pause safely. The AI suggests a test world and a few failure cases. Priya runs them and catches a timing bug that would have stopped the line later.

At the end of the day, she uses the Document and Communicate prompt to write a short summary for her team lead. The summary covers the fix, the test results, and the next check. Priya leaves on time and takes her dog for a long walk before dinner.