Present research clearly

Do not recite the poster from top left to bottom right. Begin with the problem or question, state your contribution, and use the poster when the listener needs evidence.

Build three versions from one spine

Treat these as rehearsal formats, not official presentation limits. Always follow the timing and format required by your fair, conference, classroom, or mentor.

20 seconds

Low-cost robot grippers often miss object slip when vision is blocked. I tested whether adding a simple tactile signal reduced drops in simulation. It lowered the drop rate relative to a vision-only baseline, suggesting a feasible direction for a physical prototype.

60 seconds adds the method, central number, and one limitation:

I created 120 seeded grasping trials across three object shapes and compared the same controller with and without a simulated tactile signal. The tactile version reduced drop rate from 18% to 11%; the effect was largest under partial visual occlusion. The simulation uses simplified contact physics, so the next step is testing whether the benefit survives sensor noise on a physical gripper.

Three minutes adds why you chose the problem, how the evidence was produced, why the baseline is fair, what failed, what you contributed, and what should happen next. Build this version from your real project rather than memorizing the example above.

Use your own real numbers and conditions. If results are preliminary, say so.

Answer the question beneath the question

Judges, professors, and reviewers are usually testing your reasoning and ownership.

• Why did you choose this? Connect the problem to an observation, literature gap, or user need—not “I have always loved science.”
• What was your contribution? Name what you designed, coded, collected, analyzed, or concluded.
• What is your baseline? Explain why it is a fair comparison.
• What surprised you? Describe a result that changed your model of the problem.
• What failed? Show the failure, diagnosis, and resulting decision.
• What are your limitations? Name the boundary and the test that would address it.
• What would you do next? Choose one high-information next experiment, not ten features.
• What did your mentor do? Separate their training, equipment, data, code, and decisions from yours precisely.

If you do not know, say what you do know and how you would find out. Inventing an answer damages credibility more than a bounded “I don’t know yet.”

Make mentorship transparent

Prepare a simple ownership statement:

My mentor suggested the original paper and trained me on the imaging system. I designed the comparison, wrote the analysis code, collected the final measurements, and interpreted the result. The lab supplied equipment and reviewed my safety procedure.

Adapt it to reality. Do not minimize legitimate help, and do not let group work make your contribution impossible to identify.

Practice for conversation, not theater

Record one attempt to identify unclear wording and filler. Then practice with interruptions. Ask a peer to point at any figure and say, “What am I looking at?” Ask a teacher to challenge the baseline and sample. Practice returning to the central question after a tangent.

Maintain eye contact when culturally and personally appropriate, face the listener, and pause before answering. Accessibility and clear reasoning matter more than performance tricks.