Merge pull request #266 from HIRO-group/kaleb_wiki_update

added blog post draft re: area exam

Author: alecive

_posts/wiki/2025-02-14-area-exam.md

@@ -0,0 +1,124 @@
+---
+title: Area Exam Guide
+description: A guide to passing the area exam for PhD students in the CS Department
+author: Alessandro Roncone and Kaleb Bishop
+permalink: area_exam_guide.html
+category: [wiki]
+subcategory: news
+---
+
+# A Roadmap to the Area Exam
+
+The Area Exam is a **critical milestone** in your doctoral journey. Its primary purpose is to demonstrate your understanding of the state-of-the-art in your chosen research area and to prove that you have the technical depth required to contribute to that field.
+
+This document outlines the requirements and expectations for students in the Robotics team of the Computer Science Department at University of Colorado Boulder. Unlike the flexibility of early graduate work, this exam has strict standards regarding timeline, format, and content.
+
+## 1. Picking a topic area
+Area exams are easy to postpone, because as PhD students it's easy to feel like there's always something more urgent to be working on. 
+However, _the reading and synthesis you do for this exam should provide the forward momentum for your thesis_. You should have a clear idea of what you want your research to focus on and why.
+
+**Your topic should be specific**. You cannot cover the entirety of "Human-Robot Interaction" or "Machine Learning" in 30 minutes. Work with your advisor to narrow down your methodological or application area.
+
+### Examples of Past Topic Areas
+
+To give you a sense of scope, here are examples of successful area exam topics from the HIRO and CAIRO labs:
+
+- Yi-Shiuan Tung: Generalization in Reinforcement Learning via Environment Design
+- Nataliya Nechyporenko: Contact-rich Planning for Robotic Manipulation
+- Caleb Escobedo: Reactive Robot Control and Sensing Systems
+- Aaquib Tabrez: Explainable AI, Trust, and Influence  
+- Matthew Luebbers: Shared Mental Modeling  
+- Carl Mueller: Constrained Motion Planning
+
+## 2. Setting a Timeline & Requirements
+
+The logistics of the exam are as important as the content. **You are responsible for managing this process.**
+
+### The 30-Minute Standard
+
+The presentation must be strictly 30 minutes long. This is the new standard agreed upon by the robotics faculty (including Profs. Correll, Heckman, and Hayes). You will be cut off if you go significantly over time. **Being concise is a requirement:** it is harder to effectively convey complex content in a limited amount of time than having any time in the world. 
+
+### Mandatory Deadlines
+
+To ensure you pass, we enforce the following internal deadlines. Failure to meet these may result in your exam being postponed.
+
+ * **4+ weeks out:**  
+   * _Finalizing the Committee_: Talk to your advisor to finalize your topic and identify two other faculty members. Send a brief email describing your topic and why their expertise is relevant.
+   * _Finalizing the list of papers_: Finalize your reading list with your advisor.
+
+ * **2 weeks out:**  
+   * _Draft of Area Exam Document_: You must send your advisor an almost-complete draft of the area exam document. This should be a 2–4 page review or survey of your chosen area, written in LaTeX using the IEEE two-column format. This document will serve as the spine around which the presentation is created.
+
+* **1 week out:**  
+   * _Finalize Area Exam Document_: Incorporate your advisor’s feedback and bring the document to a polished, near-final state. This document should be ready to share with your committee and should closely mirror the structure of your planned presentation. This document should then be sent to your committee with enough time for them to review it before the exam.
+   * _Draft of Area Exam Presentation_: You must send your advisor your complete slide deck one week before the exam. This will be enforced. If there is no draft by this time, the exam will not proceed.
+   * _Practice Talk_: You must schedule a practice talk with the lab at least one week before your actual exam. This allows time for major restructuring if necessary. Ask for constructive but harsh feedback from senior Ph.D. students that have already passed the exam.
+
+## 3. Structure of the Talk
+
+Your 30-minute presentation should not be a flat list of paper summaries. **It needs a narrative arc.** We expect the talk to be roughly split in half:
+
+### Part 1: Breadth (The Landscape) ~20 mins
+
+- **Intro & Background:** Define the problem space. Why does this area matter?  
+- **Bibliography Review:** Synthesize the "state of the art." Do not just list papers; categorize them (for example, "Approaches to State Estimation" or "User Modeling Techniques").  
+- **Taxonomy:** Show how you organize the field mentally.
+
+### Part 2: Depth (The Deep Dive) ~10 mins
+
+- **The Deep Dive:** Pick one single paper and go deep. Explain the math, the algorithm, or the specific technical contribution in detail. You must show that you understand the mechanics, not just the abstract. It is recommended that you choose a technical paper for your deep dive.
+- **Your Trajectory:** Conclude with your research vision. Where are the gaps? What are you going to build or solve based on what you just presented?
+
+## 3. Choosing Your Papers
+
+For each paper, you must clearly understand the contributions and why they matter.
+
+- **Relevance:** Choose papers that highlight the expertise you are developing.  
+- **Age of Papers:** There is no blanket rule against older papers. Foundational papers (even from decades ago) are often crucial for understanding the root of a field. If a 10-year-old paper is the seminal work in your area, you should include it. Use judgment and context.  
+- **Committee Work:** Do not avoid papers authored by your committee members. In fact, this is encouraged. Building upon the prior work of your advisor or committee members demonstrates that you understand the intellectual lineage of your lab and department. It is often valuable to show how your proposed work extends their findings.
+
+## 4. Making Your Slides
+
+- **Visuals over Text:** Minimize on-slide text. You want the committee listening to you, not reading.  
+- **Evidence:** If you include a figure, data table, or equation, you must explain it. Do not use complex math as "set dressing."  
+- **Synthesis:** Your slides should explicitly state the key takeaway of each section. Why should the audience care?
+
+## 5. Giving Your Talk
+
+- **Presenter View:** It is acceptable to use notes, but do not read from them continuously.  
+- **Q&A:** The closed Q&A session will test your ability to articulate opinions and defend your decision-making. You should be ready to explain what is exciting about a paper, what is missing, and how you would fix it.
+
+---
+
+# Appendix: A Student's Experience (Case Study: Kaleb Bishop)
+
+The following is a perspective from Kaleb Bishop, who successfully passed their area exam in Spring 2025.
+
+## My Topic and Approach
+
+My topic was *AI for the facilitation of task handover*. Because this is a specific application domain rather than a standard technique, my talk aimed to look at the state of the various component research areas that would be applied to this problem space.
+
+I taxonomized the component parts of the problem this way:
+
+- Forming a state estimate from input data  
+- Acquiring a listener model for producing model updates  
+- Performing model updates given a difference in mental models  
+
+## My Paper List
+
+My advisors and I settled on a list that included:
+
+- Zhang et al. (2022) on multimodal summarization (for state estimation)  
+- Das, Chernova & Kim (NeurIPS 2023) on State2Explanation  
+- Zhang & Soh on LLMs as zero-shot human models  
+- Chakraborti et al. (2017) on Model Reconciliation (a foundational paper in the field)  
+
+## The Outcome
+
+I passed! The committee’s questions tested my ability to articulate my opinions: what is or is not exciting about a particular paper, and what the biggest challenges are in the area.
+
+## The Aftermath
+
+You can view the slides from my talk [here](http://kalebishop.github.io/files/Bishop_prelim.pdf). If you're a CU affiliate you can view a recording of it [here](https://o365coloradoedu-my.sharepoint.com/:v:/r/personal/brha2830_colorado_edu/Documents/CAIRO%20Lab/Talks/Prelim%20Exam%20Talks/2025%20Bishop/Bishop_areaexam_recording.mp4?csf=1&web=1&e=Odtabv).
+
+I'm so grateful to have gotten through this process with my pride mostly intact. I'm also very thankful to my committee for their patience with me---it took me a long time and a lot of iteration to get the presentation to a state we all were happy with. I wrote this post in hopes that it would help make the process a bit less intimidating for future students. Good luck!