Robotics Lab becomes an engineering performance environment where learners make predictions, act, revise, transfer, and produce evidence.
An engineering bay with constraints, telemetry, safety margins, iteration history, and failure recovery. This V10 profile is intentionally separate from the current Robotics Lab runtime until the new shell is built and validated.
V10 forbids blank-canvas confusion: the first action must be visible, meaningful, and produce a consequence.
The learner sees one mission goal before any controls compete for attention.
The first action is obvious, clickable/touchable, and visibly changes state.
The mission requires at least three choices where different reasoning produces different outcomes.
The sim includes a likely misconception and captures evidence when the learner falls into or avoids it.
The learner can revise after observing consequences instead of being graded only on first output.
The learner applies the same reasoning to a new context before completion.
TeacherOS can summarize prediction, action, observation, explanation, revision, and transfer.
The sim must resist cheating, reduce burnout, reduce grading time, maximize screen-time value, and support misconception-cluster feedback.
Teacher-provided sources, datasets, documents, and research constraints must be preserved in the mission contract.
TeacherOS/TeachProof must recommend how the classroom simulation should have been run based on observed student events.
The selected shell changes language density, controls, read-aloud, and scaffolding by grade band.
The experience is Chromebook-safe, keyboard reachable, reduced-motion aware, and readable.
The sim is not a reskin: objects, controls, mission art, and success state are domain-specific.
These are the original STEM missions upgraded into the V10 workflow: start, goal, controls, evidence, submit.
Click the move/help control
Save a prediction about whether grip, path, or speed will cause the failure