College Timetable Auto-Scheduling System with Constraint-Based Slot Allocation

Every semester, department coordinators lose days — sometimes weeks — to hand-built timetables: matching faculty availability against room capacity, keeping lab sessions in contiguous blocks, and redoing the entire grid whenever one input changes. Manual grids routinely hide clashes that surface only after classes begin: a faculty member in two rooms at once, or a lab split across non-adjacent periods. Underneath, timetable construction is a classic constraint-satisfaction problem — sessions must be assigned to (day, period, room) triples so that a set of hard constraints holds and soft preferences are optimised. This project implements that formulation as a real working system: a web app where the department enters faculty, subjects, rooms and the weekly slot structure, and a solver produces a verified timetable with a conflict report and CSV/PDF export. Because the constraint model and the solving algorithm are separate, the student can explain exactly what "valid" means and exactly how the timetable was found — the kind of depth that carries a viva.

1. The department enters faculty (teachable subjects, max periods per day, blocked slots), subjects (weekly period requirement, lab flag), rooms (classroom or lab, capacity) and the weekly slot structure.
2. Subjects are expanded into sessions: lab subjects become double-period contiguous blocks, theory subjects become single periods; sessions are ordered by constrainedness (labs first, then fewest faculty options).
3. The shipped build encodes the model in Google OR-Tools CP-SAT; the web demo runs the same constraint set with greedy seeding followed by chronological backtracking search.
4. Placement tries day, period and room assignments, rejecting any that violate an enabled hard constraint, until every session is placed or the search is exhausted.
5. Soft constraints are scored on the result — slot preferences, load balance against the weekly target, and empty gaps between classes.
6. An independent post-solve scan re-validates the final grid against every hard constraint; unplaceable sessions are listed in the conflict report with likely causes.
7. The verified timetable renders as a weekly grid with CSV export and print/PDF output for departmental circulation.

Tech stack:

• Python 3, Google OR-Tools (CP-SAT) — shipped solver backend
• Flask backend, SQLite (PostgreSQL on request)
• HTML, CSS, JavaScript — frontend and the greedy + backtracking demo solver
• CSV export, print/PDF via browser

• [Hard-constraint solver] Every generated timetable respects all enabled hard constraints by construction: no faculty double-booking, no room clash, lab sessions in contiguous blocks, daily period caps per faculty, and lab/classroom room-type matching.
• [Soft-constraint optimisation] Faculty slot preferences, balanced teaching loads and compact days (minimal gaps) are scored after each run — the report shows a soft-constraint score out of 100 with reasons for every deduction.
• [Conflict report] When the inputs are infeasible, the system lists exactly which sessions could not be placed and the likely cause (overloaded faculty, insufficient lab windows, demand exceeding capacity) instead of silently producing a broken timetable.
• [Weekly grid + faculty load views] Colour-coded timetable grid, per-faculty weekly load bars with an overload flag above the department target, and a solver log that traces the search for transparency.
• [Editable department model] Add faculty with teachable subjects, daily caps and blocked slots; subjects with weekly periods and lab flags; rooms with type and capacity — then re-solve in one click.
• [CSV and print/PDF export] One-click CSV download of the solved timetable and a print stylesheet for notice-board circulation.
• [Multi-department solving (optional)] Extends the model to shared faculty and room pools across departments, solved jointly instead of per-division.
• [Elective clash handling (future scope)] Student-choice electives with student-level clash avoidance — documented as a roadmap item, not shipped in the base build.

• Complete solver backend source code (Python + Google OR-Tools CP-SAT)
• Web application: department data management, one-click solve, weekly grid, faculty load dashboard, constraint toggles, conflict report and solver log
• Working JavaScript demo with realistic sample data (Computer Engineering semester: 6 subjects + 2 labs, 5 faculty, 4 rooms)
• CSV export and print/PDF output
• Admin workflow documentation (how a coordinator enters data and runs the scheduler)
• Project report PDF (constraint modelling, solver design, methodology, results)
• PPT presentation for final review
• Viva Q&A preparation document (CSP formulation, CP-SAT, backtracking, hard vs soft constraints)