AI for Electrical and MEP Design Engineers

It is a free-to-learn online course that teaches electrical and MEP design engineers to use AI assistants such as ChatGPT, Claude, Gemini and Microsoft Copilot on real design work, with every result checked by an independent route. It covers calculation prompts, load and cable sizing checks, lighting and chilled water checks, specifications, method statements, BOQ and take-off, code look-up, Python scripts, PDF and drawing extraction, QA of AI output and client confidentiality.

It can draft and run them, but every answer must be checked by a second route before it is used. Assistants predict text, so they make specific, confident errors such as dropping the root 3 in a three-phase current, using 20 C cable resistance for a hot cable or treating kW as per hour. The course shows how to give the inputs, make the arithmetic run in code and verify the result by hand, spreadsheet or a tested script.

No AI output is compliant until a competent engineer has checked it against the adopted edition of the code. Assistants rarely see the full licensed text, mix editions and drop exceptions, for example quoting Article 220 for load calculations that moved to Article 120 in the 2026 NEC. The course teaches look-up with editions and citations, grounding in licensed documents and verification in the code itself.

Only if the client contract and your company's AI policy allow it, and then only on an approved plan with the data classified and redacted first. Consumer plans may use chats for training unless you switch it off, while business plans of the major assistants exclude business data from training by default (as of October 2026). The course covers the settings, a four-level data classification and redaction of drawings and schedules.

It is for electrical, mechanical and building services design engineers, MEP coordinators, estimators and graduate engineers in consultancies and contractors worldwide, including the Gulf. You should know what a load schedule, a voltage drop check and a specification are; no coding experience is needed. Electrical Design Engineering is the companion course for the calculations themselves.

One AI assistant is enough: ChatGPT, Claude, Gemini or Microsoft Copilot, and the free tiers work for most exercises. You also need a spreadsheet such as LibreOffice Calc, and Python with pandas and pdfplumber in Google Colab for two modules, with nothing to install. Product and model names in the course are dated October 2026 because they change often.

About 18 hours at your own pace, of which just under 11 hours is video, plus about 12 hours for the optional project. Every module, the notes, worked problems, practice tasks, project and final assessment are free to learn.

You get a verifiable certificate of completion from EDWartens when you finish the modules and pass the 15-question final at 60 percent. It carries a unique certificate number and a QR code that open a public verification page showing the course, the modules covered and your final score. It is not a professional engineering licence or a vendor certification from OpenAI, Anthropic, Google or Microsoft.