PadhaikoAll subjects

Progress Report Submission and Presentation II

Project II (ENCT 461)

Open in the Padhaiko app

4 Progress Report Submission and Presentation II

Progress Report Submission and Presentation II is the mid-semester evaluation gate held 2 months after the commencement of the eighth semester. At roughly the halfway mark between Project II start and Final Defense, this checkpoint tests whether your group has moved decisively beyond prototypes into an integrated, testable system or active experimental evaluation.

Definition

Progress Report Submission and Presentation II (Month-2 Checkpoint)

Under the ENCT 461 syllabus, Progress Report Submission and Presentation II takes place two months after the eighth semester begins. It serves as the mandatory midterm verification that core software/hardware modules are integrated or primary experimental runs are producing preliminary quantitative results, leaving the remaining weeks strictly for systematic validation, optimization, user testing, and final report preparation.

4.1 Role and Timing of the Month-2 Checkpoint

Consider the official ENCT 461 calendar:

Because only three to five working weeks separate Month 2 from Final Report submission, Presentation II functions as a practical feature-freeze boundary:

  1. Product Track: All primary modules (database, backend APIs, hardware sensing/control, core UI flows) should be integrated end-to-end so the final month goes to bug fixing, load/security testing, UAT with target users, and report writing.
  2. Research Track: Data preprocessing, baseline runs, and main candidate experiments must be running at Month 2 so the final month focuses on ablations, error analysis, statistical comparison, and thesis drafting.
  3. Internship-Based Track: The host module must be running on anonymized/sanitized demo inputs with measurable metrics verified by both the industry mentor and department supervisor.
AspectPresentation I (Week 2)Presentation II (Month 2)
Core Question"Is your refined Sem-8 plan solid and ready to build?""Does your integrated system/pipeline actually work end-to-end?"
Expected EvidenceProject I feedback matrix, updated diagrams, environment/dataset readinessLive integrated demo, preliminary test logs, early benchmark tables/plots
Scope DecisionsClarify and lock refined deliverablesFeature freeze: cut unbuilt nice-to-have extras and lock evaluation criteria
Panel ScrutinyArchitecture soundness and timeline realismBugs, edge cases, metric validity, individual member coding depth

4.2 Contents of Progress Report II

Progress Report II (typically 15 to 25 pages in the department format) updates the supervisor and committee with hard engineering evidence rather than planned intentions.

Essential Sections of Progress Report II
SectionConcrete Content Expected at Month 2What Evaluators Check
1. Milestone SummaryPlanned vs. completed tasks since Presentation I (Week 2 to Month 2)Honest percentage completion backed by implemented modules
2. Implemented Architecture / ModulesCode/hardware modules built, finalized DB/API schemas, hardware wiring/PCB photosAlignment between diagrams and actual running codebase
3. Preliminary Testing & ResultsUnit/integration test logs, preliminary latency/accuracy tables, sample outputsReal numbers and failure cases (not fabricated 100% pass tables)
4. Deviations & BottlenecksTechnical blockers encountered (e.g., sensor drift, class imbalance) and fixes appliedEngineering maturity in diagnosing and resolving issues
5. Post-Month-2 Completion PlanExact tasks for Weeks 9 to Final Report (UAT, ablations, tuning, report chapters)Feasibility of finishing before the End - 3 weeks report deadline

4.3 Live Demonstration and Presentation II Strategy

Unlike Presentation I, where slides dominate, Presentation II expects a live working demonstration (or live terminal/UI walkthrough of the research pipeline).

  1. Demonstrate the real end-to-end flow: Show an actual input entering the system (sensor reading, user form submission, or test video/text sample), passing through your backend/model, persisting or computing a result, and displaying the output cleanly.
  2. Prepare a disaster-proof backup: Conference-room Wi-Fi drops, USB cables loosen, and cloud free tiers sleep. Always record a crisp, unedited 2-minute backup screen/camera video of the live run beforehand—use it only if hardware or network fails in the room.
  3. Present imperfect preliminary results honestly: If your current model gets 0.76 macro-F1 on a hard class or your API latency spikes under 50 concurrent users, present the confusion matrix or latency trace alongside your exact plan for Weeks 9–11 to improve it.
  4. Confirm individual contribution readiness: Every group member should demo or explain the module they personally implemented and answer technical questions without leaning on one spokesperson.
Example 4.1

Handling an Incomplete Module Honestly at Presentation II

Two weeks before Presentation II (Month 2), a three-member Product-Track group building a Smart Hostel Energy & Fault Monitoring System has completed ESP32 current-sensing firmware, MQTT ingestion, PostgreSQL storage, and the live consumption dashboard—but their automated tariff-prediction module is unfinished and buggy. How should they handle Progress Report II and Presentation II?

Audit what is solidly complete vs. unstable: Core metering, real-time MQTT telemetry, threshold over-current alerts, and historical daily/weekly charts are tested and stable. The tariff-prediction regression model is still overfitting on two weeks of noisy readings.

Stabilize the demo branch and separate experimental work:

Keep main clean with the rock-solid live metering, over-current alert workflow, and historical analytics. Include unit/integration test logs for MQTT ingestion and threshold alerts in Progress Report II.

Report the unfinished module transparently with a realistic Week 9–10 plan: In Progress Report II and Presentation II, demo the live ESP32-to-dashboard pipeline first, then show a single slide on the tariff-prediction status: explain why two weeks of data caused overfitting, present initial MAE numbers, and ask the supervisor whether to collect two more weeks of hostel meter logs or narrow prediction to simple moving-average forecasting before the post-Month-2 feature freeze.

Answer:

The group stabilizes the working core pipeline on main for a reliable live demo, documents completed telemetry test logs in Progress Report II, and presents the tariff-prediction bottleneck honestly with preliminary MAE numbers and a supervisor-guided decision to refine or scope it down after Presentation II.

Example 4.2

Structuring a Month-2 Live Demo and Slide Deck for a Research-Track Project

A two-member Research-Track group working on Nepali OCR Post-Correction using Sequence-to-Sequence Models is preparing for Presentation II (Month 2). They have baseline Tesseract OCR results and preliminary fine-tuned mT5-small outputs on 3,000 scanned textbook lines. Outline their Presentation II deck and live demo plan.

Keep background recap under 90 seconds and jump straight to system & data status:

  • Slide 1–2: Problem statement recap + dataset statistics (3,000 verified Devanagari textbook line images, train/val/test split by book chapter so zero page leakage occurs).
  • Slide 3: End-to-end OCR + post-correction pipeline architecture.

Present preliminary comparative table and error taxonomy:

  • Slide 4–5: Preliminary Character Error Rate (CER) and Word Error Rate (WER) comparing Raw Tesseract vs. Rule-Based Spellcheck vs. Fine-Tuned mT5-small on the validation split.
  • Slide 6: Error breakdown showing where mT5-small helps (broken matras, conjunct consonants) and where it still hallucinates (proper nouns and numerical tables).

Run a controlled live inference demo plus Month-3 plan:

  • Live Demo (2 minutes): Feed 3 unseen scanned Devanagari line images into the CLI/web inspector showing Raw OCR -> Corrected Output -> Ground Truth -> CER delta (with a pre-recorded screencast backup ready).
  • Final Slides: Post-Month-2 plan (Weeks 9–11) covering constrained decoding/lexicon guard against number hallucination, inference speed benchmarking, and drafting Chapters 4–5.
Answer:

The deck skips lengthy intro material to highlight leak-free dataset splits, preliminary CER/WER comparisons, an honest error taxonomy (matra fixes vs. numeric hallucinations), a 2-minute live unseen-sample demo with video backup, and a concrete Weeks 9–11 hardening plan.

Viva-Style Questions: Progress Report Submission and Presentation II

Question 1

When is Progress Report Submission and Presentation II held in ENCT 461, and why is Month 2 considered the natural "feature-freeze" point?

Official timing: The syllabus schedules Progress Report Submission and Presentation II 2 months after the commencement of the eighth semester.

Why feature freeze follows Month 2: Final Project Report submission occurs 3 weeks before semester end, leaving barely a month after Presentation II. Freezing new features right after Month 2 ensures sufficient time for systematic unit/integration/user testing, performance optimization, comparative benchmarking, and writing Chapters 1–6 cleanly.

Answer:

It occurs 2 months into the eighth semester. Because Final Report submission is due 3 weeks before semester end, freezing new features at Month 2 protects the remaining weeks for thorough testing, evaluation, and thesis writing.

Question 2

How should Progress Report II differ from Progress Report I in content and tone?

Shift from plans to verified evidence: Progress Report I (Week 2) focuses on Project I feedback resolution, refined architecture, and upcoming work plans. Progress Report II (Month 2) documents implemented code/hardware, working integration, preliminary test and benchmark outputs, and resolved vs. open technical blockers.

Remaining schedule precision: The forward schedule in Report II shifts from broad development sprints to a tight Week 9–12 checklist focused specifically on UAT, ablation/tuning, and report chapter milestones.

Answer:

Progress Report I is a refined Week-2 execution plan, whereas Progress Report II is an evidence-backed midterm audit containing implemented modules, preliminary test/benchmark tables, resolved technical bottlenecks, and a tight final-month validation/writing plan.

Question 3

During Presentation II, your live hardware or cloud demo fails to connect to the campus Wi-Fi. What preparation prevents this from ruining the evaluation?

Pre-configured local fallback and backup video:

  1. Configure a local offline mode (such as a mobile hotspot, localhost Docker stack, or pre-seeded local test database) so the demo does not depend solely on unreliable room Wi-Fi.
  2. Always bring an unedited 2-minute screen/camera recording captured the night before, showing the exact working flow with timestamps.

Professional communication with the panel: Briefly explain the network constraint, switch immediately to local execution or the backup video without wasting five minutes troubleshooting Wi-Fi settings in front of evaluators, and invite them to inspect the code/logs afterward.

Answer:

Run a localhost/hotspot offline demo mode whenever possible and always keep a fresh, unedited backup video of the live run so Wi-Fi or cloud outages never derail the presentation.

Question 4

Why do evaluators at Presentation II appreciate seeing preliminary error analysis or failed test cases alongside passing results?

Indicator of genuine testing: Real systems rarely achieve 100% accuracy or zero bugs on first integration. Showing specific edge cases that failed (e.g., blurred input images, concurrent booking collisions, noisy sensor spikes) proves the team actually ran rigorous tests rather than fabricating numbers.

Demonstrates engineering problem-solving: Explaining why a failure occurred and what fix is scheduled for Weeks 9–10 turns a weakness into evidence of systematic engineering evaluation.

Answer:

Honest preliminary failure cases prove that real testing is occurring and allow evaluators to give targeted mid-course guidance while there is still time to fix root causes before final submission.

Question 5

If evaluators at Presentation II suggest adding a large new feature that would take four weeks to build, how should the group respond?

Assess impact against the Week N-3 report deadline: With Final Report submission due 3 weeks before semester end, taking on a four-week new feature risks leaving core modules untested and the report incomplete.

Negotiate a bounded scope with the supervisor: Clarify politely during Q&A what is already in the approved Project I scope, note the suggestion as a candidate extension, and meet your assigned supervisor immediately after the session to decide whether to implement a minimal lightweight version or formally place the large addition into Chapter 6: Future Work while prioritizing thorough validation of existing features.

Answer:

Evaluate feasibility against the End - 3 weeks report deadline, consult your supervisor immediately after the review, and either implement a minimal bounded version or document the larger feature in Future Work rather than jeopardizing core testing and report quality.

← Implementation and TestingFinal Project Report Submission →