
AI can speed up exploration, code and documentation. Your team still owns every engineering decision and result.
Ask AI for design alternatives, then compare them against requirements, constraints, risks and trade-offs before choosing.
Re-derive critical steps, check dimensions, inspect assumptions and test boundary cases. A confident AI answer can still be wrong.
Run unit tests, compare with a baseline, inspect edge cases and reproduce the output. Working once is not validation.
Verify components, ratings, tolerances and safety rules in official datasheets and applicable standards—not an AI summary.
Record prompts, useful outputs, rejected suggestions, human edits, tests and final decisions so your process remains explainable.
Follow your university's policy. Do not upload confidential research, sponsor data, personal information or restricted designs without permission.
Be ready to explain why the design works, how you tested it, where it can fail and why you rejected the alternatives.
ABET's 2026–27 criteria still center design, standards, constraints, experimentation, communication, ethics and engineering judgment.
Read the Engineering AI Guide