[HD] EECS3342 F26 - 2026-09-10 (Thursday) - Lecture 1

Overview

Jackie Wang introduces EECS 3342, explaining its assessment policies, study resources, and emphasis on learning formal specification and verification through independent reasoning. The course moves from a discrete-math review into Event-B and Rodin modeling, using a bridge controller to develop finite-state models, invariants, progress properties, and refinements; Wang also explains how to use AI as a learning aid without outsourcing the reasoning students must demonstrate.

Key takeaways

The assessment structure prioritizes demonstrated understanding: four timely lab submissions earn 5% total, written tests are worth 15% each, and the cumulative paper-and-pen final accounts for 50%.
The October 7 programming test covers Labs 1 and 2 and uses Rodin, so students should learn the Event-B tool and its modeling conventions while completing those labs.
Wang permits AI, peer help, and other resources for lab learning, but tests require independent work; AI-generated answers are not a substitute for the reasoning assessed later.
The course’s central technical progression is from discrete-math foundations to finite-state Event-B models, then to invariants, progress properties, and refinement through a bridge-controller case study.
Formal reasoning helps engineers check whether specifications are complete and unambiguous and whether system outputs satisfy them—judgments that remain necessary when AI generates code or other outputs.

Chapters

0:00 EECS 3342: Course Purpose and Formal Verification

Jackie Wang describes EECS 3342 as a challenging introduction to formal verification and system specification.
The first lecture focuses on the syllabus and course rationale; technical instruction begins the following week.
Wang plans to teach examples from scratch so students learn the reasoning process needed for cumulative assessments.

2:03 Communication, Course Access, and Optional Bonus Surveys

Students may call Wang Jackie and can seek help during in-person office hours or after class.
Two anonymous surveys may provide bonus credit if participation reaches a sufficient threshold: a midterm course evaluation and an end-of-term lecture-digest survey.
Students not yet enrolled should email their name, student number, and York ID to request access to eClass, especially for lab instructions.

4:43 Lecture Expectations and Learning from Worked Examples

Lecture attendance is encouraged but not required because recordings are available; Wang asks in-person students not to talk, use phones distractingly, or work on unrelated tasks.
Laptops are welcome for slides and lecture notes, while headphones and unrelated computer use may lead to a warning or a request to leave.
Lectures will work through examples from scratch rather than simply read slides, supporting the independent reasoning needed for the cumulative exam.

7:41 Two Open Letters on Foundations and Formal Verification

Wang assigns two short open letters for weekend reading: one on foundational learning and course expectations, and one on why formal verification matters in the age of AI.
The letters explain the test-focused assessment structure and why the course emphasizes independent reasoning rather than projects.
Wang recommends using AI interactively to support learning, but cautions that generated solutions may be inelegant and should not replace students’ own thinking.

9:56 eClass, Course Materials, and Event-B Textbook Access

Students should check eClass announcements regularly for updated study materials, test guidance, and practice questions.
Wang says his slides and annotated notes are sufficient as the required study materials; there is no textbook to purchase.
A permitted draft of Modeling in Event-B is available through eClass for York University students, but students must not share its PDF publicly.

13:05 Lecture Website: Notes, Digests, and Course Calendar

Wang’s public lecture website hosts recordings, slides, notes templates, annotated notes, and a lecture digest for each class.
Notes templates are intended to appear at least one day before class; the digest icon links to a study roadmap for that lecture.
Wang recommends bookmarking the course page and saving the semester calendar, which lists milestones and assessment dates.

16:22 Math Review Leads into the Bridge-Controller Case Study

The course begins with several classes reviewing logic and discrete mathematics, including quantifiers, implications, sets, relations, and partial and total functions.
After the reading week, students begin a bridge-controller case study that applies refinement to a reactive, safety-critical system.
The study sequence prepares students for both the first programming test and the first written test.

23:26 Lab Credit, Test Rules, and Final-Exam Format

There are four labs worth 5% total; submitting a best attempt by the deadline earns the lab credit even if the work is incomplete.
Students may use peers, online resources, and generative AI for labs, but Wang urges them not to submit an AI-generated solution without working through the reasoning.
Programming and written tests must be taken independently in the student’s enrolled lab section, with no internet, AI, or slides; the three-hour cumulative final is paper-and-pen and allows a data sheet.

36:08 Assessment Weights: Labs, Tests, and the 50% Final

Four labs contribute 5% total, with each lab worth 1.25% for a timely submission.
One Rodin programming test covers Labs 1 and 2; two written tests are worth 15% each and focus primarily on multiple-choice questions with possible short typed responses.
The final exam is worth 50% and is cumulative for material taught after the introductory lecture; Wang expects short answers, proofs, justifications, and model development rather than mostly multiple choice.

40:29 Grade Cutoffs, Attendance, and iClicker Setup

Grades are calculated from weighted marks using standard cutoffs, but Wang may adjust letter-grade thresholds at semester’s end; he does not promise a conventional curve.
Attendance is not a graded requirement or guaranteed bonus, but Wang may consider it when deciding whether to raise a borderline letter grade.
Students should install iClicker and enroll in EECS 3342; trial attendance checks are planned for the next Tuesday and Thursday lectures.

46:31 Key Dates: October 7 Programming Test and December Exam Window

Lab 1 is due about two weeks after release, and Wang recommends starting early because Lab 2 will also be assigned with a two-week completion window.
The programming test is scheduled for October 7 during students’ enrolled lab sessions and covers Labs 1 and 2.
Written Test 1 is tentatively scoped through Lecture 11; Written Test 2 is planned later in the term, and the final exam is scheduled sometime between December 10 and 23.

55:38 Why Formal Reasoning Still Matters in the Age of AI

Wang argues that students need formal foundations to judge whether an AI prompt is complete, unambiguous, and correct.
Testing AI output is useful, but proofs and formal arguments can provide another way to assess whether a result meets the intended specification.
The course emphasizes abstraction, logic, and independent reasoning so future software engineers can exercise professional judgment rather than rely entirely on AI.

1:00:09 Open-Letter Guidance: Active Study and Responsible AI Use

Wang’s first open letter uses the logic of implication and its contrapositive to distinguish teaching quality from students’ own controllable study effort.
He recommends active engagement and mastery-oriented learning, and advises students to use sample questions as a diagnostic rather than as a substitute for studying course materials.
The second letter explains the value of formal verification and Event-B even if students do not use that specific tool after the course; Wang encourages students to seek help early if they struggle.

1:09:04 Course Outcomes: Requirements, Invariants, and Finite-State Models

Students will distinguish environment descriptions from requirements descriptions and express safety constraints as invariants.
The bridge controller illustrates progress requirements: a deadlocked system cannot proceed, while a livelocked system remains active without making useful progress.
Labs and the case study develop finite-state machine models using constants, variables, events, axioms, invariants, and guards.

1:12:30 Sets, Predicate Logic, Refinement, and Proofs

The course reviews set theory and predicate logic, including membership, subset versus proper subset, and universal versus existential quantification.
Refinement is introduced informally as building an improved model from an earlier one; later work makes the refinement conditions precise enough to prove.
Students will practice sequence-style proofs and apply formal modeling to the bridge-controller reactive system.

Summary, takeaways, and chapters were generated by AI from the video's transcript and may contain errors. The video belongs to its creator, Jackie Wang.