A high school student can do publishable engineering or robotics research by building or simulating a system that answers one measurable question, then comparing it against a clear baseline. The difference between a cool build and a research paper is the test: what you measured, what you compared it with, and whether the result holds up.
This guide covers how to turn an engineering interest into a research question, the methods a student can carry out with low-cost hardware or free simulation tools, where student engineering papers get published, which competitions fit, and the engineering and robotics papers YRI Fellows have written.
Building versus researching
Many students who love engineering already build things: robots, drones, apps, devices. Admissions readers and judges see a lot of builds. What they see far less often is an engineering study.
| A build | A research project |
|---|---|
| "I made a robotic arm" | "Which classifier controls a low-cost arm most accurately from muscle signals?" |
| "I built a drone" | "Can a decision layer pick safer emergency actions than a fixed rule?" |
| "I designed a helmet" | "Does fitting a liner to scan data change how well it fits the head?" |
| Works or does not work | Measured against a baseline, with numbers |
You can keep the build. The research is the question you wrap around it.
Areas of engineering research for students
| Area | Example focus | What you need |
|---|---|---|
| Robotics and control | Arms, grippers, mobile robots, gesture control | Microcontroller, sensors, motors |
| Aerospace and drones | Flight safety, path planning, materials | Flight simulators, public materials data |
| Embedded systems and IoT | Sensors, signal processing, low-power devices | Arduino, ESP32 or Raspberry Pi |
| Biomedical engineering | Prosthetics, wearables, protective gear | Sensors, 3D scanning or printing, public scan data |
| Energy | Harvesting, storage, efficiency | Simulation, simple test rigs |
| Environmental engineering | Water treatment, pollution monitoring | Public water quality data, models |
| Manufacturing | Machining, optimization, digital twins | Simulation, public process data |
Example research questions
Good engineering questions compare designs or methods under defined conditions:
Robotics and control:
- Which machine learning model classifies forearm muscle signals most accurately on a $35 single-board computer?
- How much does adding a second sensor improve a mobile robot's obstacle detection in low light?
Aerospace:
- Can a drone rule out physically impossible emergency actions before choosing one, and does that reduce unsafe choices?
- Which composite materials offer the best thermal protection per unit weight, based on public materials data?
Energy and environment:
- Which vibration energy harvester design captures the most energy from everyday vibrations, not lab-perfect ones?
- Where should small treatment units sit along a polluted river to cut the most contamination?
Biomedical engineering:
- Can scan-based head geometry produce a better-fitting helmet liner than standard sizes?
Methods a student can actually carry out
1. Build a low-cost prototype and test it
Microcontrollers and single-board computers such as Arduino, ESP32 and Raspberry Pi make it possible to build working sensing and control systems for very little money. The research step is designing the test: what conditions, how many trials, which metric, and what you compare it with.
2. Simulate before or instead of building
Free and open simulation tools let a student test designs at a scale no garage could match. Flight stacks and robot simulators such as PX4 and Gazebo can run hundreds of matched trials. Simulation results are publishable when the setup is documented and the comparison is fair.
3. Run a comparative design study
Some of the cleanest engineering papers compare two or three designs under the same conditions and report which wins and why. This is often more valuable than inventing something new, because the answer helps every engineer choosing between those designs.
4. Use public engineering data and machine learning
Agencies and research groups publish materials, sensor and process data. A student can train models that predict performance and point to better designs, as one Fellow did with NASA thermal protection materials data.
5. Design a framework and evaluate it with data
Systems papers propose a design, such as a treatment network or a monitoring system, and evaluate it against measured data. The design has to be specific enough that someone else could build it.
Engineering and robotics papers by YRI Fellows
All of these are first-author papers written with YRI PhD mentors. More are on our published work page.
- Raeyaan M., an 11th grader, built a low-cost bionic arm that reads forearm muscle signals, classifies gestures with machine learning and drives servos instantly. Random Forest on raw signal envelopes reached 93% live accuracy on a $35 Raspberry Pi. Published and presented at IEEE WcCST-2026 and listed in IEEE Xplore. Read his story.
- Amaay G. built SkyShield, a decision layer that picks one of seven emergency actions for a multirotor drone after ruling out what the drone physically cannot do. He evaluated it across 4.2 million paired decisions, 209 full-flight replays and 450 matched PX4 and Gazebo simulation runs. First-author paper accepted at an IEEE conference. Read his story.
- Elliot W., a 10th grader, developed an MRI-guided approach to personalized helmet liner design using 3D head scanning. His paper was accepted at BioMRH 2026 in Glasgow, UK, with proceedings in Springer Lecture Notes in Networks and Systems. Read his story.
- Ariham J., a 10th grader, compared passive linear, symmetric bistable and tunably asymmetric vibration energy harvesters under representative everyday vibrations. First-author paper accepted at an IEEE conference. Read his story.
- Arjun V. used machine learning on NASA TPSX materials data to predict how fibre-reinforced composites perform as spacecraft heat shields, identifying LI-900 and carbon-phenolic as the strongest insulation combinations. Read his story.
- Adit A. built an adaptive digital twin for CNC machining that optimizes speed, quality and cost together. Reviewers noted it outperformed fixed, rules-based and standard machine learning strategies. Read his story.
- Sai T., a 9th grader who rated his own coding 1 out of 5, designed a decentralized in-river treatment framework for the Yamuna River with machine learning water quality prediction. Accepted at IEEE ICSCSS 2026. Read his story.
- Mohammed A., an 11th grader, studied where validation checks should sit when language models plan missions for teams of drones. First-author paper accepted at an IEEE conference. Read his story.
Where student engineering papers get published
IEEE conferences. IEEE is the largest home for engineering research, and its conferences cover robotics, control, embedded systems, energy and biomedical engineering. Accepted papers appear in IEEE Xplore, where anyone can look them up. See how to publish in IEEE as a high school student.
Springer proceedings. Many robotics, mechatronics and applied engineering conferences publish proceedings in Springer series, as with Elliot's BioMRH paper.
Elsevier and other journals. Journals take longer than conferences but suit mature work with strong results.
Patents. If your design is new and useful, a patent may sit alongside the paper. Our patent guide covers the process.
Engineering at science competitions
Regeneron ISEF has several engineering categories. Choose the one that matches what your project contributes:
| ISEF category | Fits projects on |
|---|---|
| Robotics and Intelligent Machines | Biomechanics, cognitive systems, control theory, machine learning, robot kinematics |
| Embedded Systems | Circuits, IoT, microcontrollers, sensors, signal processing |
| Engineering Technology: Statics and Dynamics | Aerospace, civil, mechanical, ground vehicles, control theory |
| Biomedical Engineering | Biomedical devices, sensors and imaging, biomechanics |
| Energy: Sustainable Materials and Design | Solar, wind and water power, energy storage, thermal design, triboelectricity |
| Environmental Engineering | Bioremediation, pollution control, waste management, water resources |
| Materials Science | Composites, biomaterials, nanomaterials, polymers, computation and theory |
For idea starters in each, see ISEF project ideas by category.
Team build competitions such as FIRST Robotics are excellent for skills and teamwork, but they are not research. A research paper built from a robotics interest adds something those competitions do not.
The Conrad Challenge rewards innovation and entrepreneurship, which suits engineering students with a product idea. See how to win the Conrad Challenge.
Common mistakes in student engineering projects
- No baseline. "It works" is not a result. "It beat the standard method by a measured margin under the same conditions" is.
- One trial. Repeat tests enough times to show the result is not luck.
- Testing only in perfect conditions. Reviewers want to see noise, wear, low battery, bad light.
- Too much build, too little writing. Leave time to write up the method so someone else could repeat it.
- Unsafe testing. High voltage, propellers and batteries need adult supervision. Follow your fair's safety rules.
Frequently asked questions
Can I do engineering research at home?
Yes. Low-cost microcontrollers, 3D printing and free simulators cover most student engineering projects. Several Fellows above worked entirely in simulation or with public data.
Does a robotics team count as research?
It builds strong skills, but a team build for a competition is not research on its own. Turning one design decision from your robot into a measured comparison can become a paper.
What should I learn first?
Basic programming in Python or C++ and how to read sensor data. You can learn control, signal processing and machine learning during the project with a mentor.
Do I need expensive equipment?
No. Raeyaan's published arm ran on a $35 Raspberry Pi, and Amaay's drone study ran in open simulation tools.
Is engineering research good for college applications?
A first-author engineering paper shows you can define a problem, design a test and defend a result, which is what engineering programs look for. It is also strong material for supplements and interviews.
How the YRI Fellowship helps you publish engineering research
The YRI Fellowship matches you 1:1 with a PhD mentor in engineering or robotics for 10 weeks of weekly sessions, about 5 to 8 hours of work a week. Your mentor helps you turn a build or an idea into a testable question, design the experiment or simulation, and analyze the results. An editing team works through every draft, and a publishing team gets the paper into IEEE, Springer or Elsevier journals and conferences, with free help until it is published.
87% of Fellows publish and 73% place at science competitions. ISEF and JSHS preparation and a mentor letter of recommendation are included. No research experience is needed, and decisions come within 48 hours.
Tuition is $2,997, or 3 payments of $999, with scholarships considered case by case. See pricing and more student stories.
Sources
- Society for Science: ISEF categories and subcategories
- IEEE Xplore: Raeyaan M., robotic arm controlled by EMG signal classification
- YRI case study: Raeyaan M.
- YRI case study: Amaay G.
- YRI case study: Elliot W.
- YRI case study: Ariham J.
- YRI case study: Arjun V.
- YRI case study: Adit A.
- YRI case study: Sai T.
- YRI case study: Mohammed A.
- YRI Published Work
- PX4 Autopilot
- Gazebo simulator
- NASA TPSX materials database
