The brief this answer responds to
For information and reference only – not to be submitted as your own work. This is a condensed model-answer extract that shows how a biomimetic robot design can be argued and tested. Every number is illustrative. Your own project needs your team’s own concepts, CAD, code and measurements, and you should follow your course rules […]
For information and reference only – not to be submitted as your own work. This is a condensed model-answer extract that shows how a biomimetic robot design can be argued and tested. Every number is illustrative. Your own project needs your team’s own concepts, CAD, code and measurements, and you should follow your course rules on outside help.
This extract follows the structure of the MEC482 biomimetic robot design challenge brief, using the water option W1 (fish-tail swimmer) as a worked example. It covers the biological principle, the weighted concept choice, a sizing check, the closed-loop controller, the fabricated parts and the 10-trial comparison. For a planning method, read how to approach the MEC482 biomimetic robot project.
Planning table
| Brief requirement | Where it appears in this extract | What the marker looks for |
|---|---|---|
| Biological principle and hypothesis | Section 1 | Organism, behaviour, physical principle and transferred feature, with credible sources |
| Concept selection | Section 2 | Weights set before scoring, a justified choice |
| Quantitative design and analysis | Sections 3 and 4 | At least five measurable requirements and two analyses |
| CAD and fabricated parts | Section 5 | Original parts, critical dimensions, tolerances |
| Closed-loop control | Section 6 | Sensor, controller, actuator, logged command, feedback and performance |
| Experiment with at least 10 trials per condition | Section 7 | Baseline, repeatable protocol, uncertainty, confounders |
1. Biological principle and hypothesis
Fish swim by passing a wave of bending along the body and tail, and many species stiffen or relax the tail to tune thrust. Vincent et al. (2006) describe biomimetics as transferring a mechanism by abstracting its principle, not copying its shape. Ijspeert (2014) shows that robots are useful for testing how such mechanisms behave, because a robot lets the engineer switch one feature on and off.
The transferred feature chosen here is a compliant tail fin whose stiffness lets it bend passively during the stroke. The hypothesis is that, at the same servo amplitude and beat frequency, a compliant fin gives a higher mean forward speed than a rigid flat-plate fin of the same area. The baseline is therefore the rigid fin on the same body.
2. Concept selection (weighted matrix)
The three water options in the brief were scored against criteria whose weights were fixed first. Scores run from 1 (weak) to 5 (strong) and are illustrative.
| Criterion | Weight | Fish-tail swimmer | Manta-ray fin robot | Duck-inspired paddler |
|---|---|---|---|---|
| Expected performance gain | 0.30 | 4 | 4 | 3 |
| Ease of fabrication in the time available | 0.25 | 4 | 3 | 3 |
| Control complexity | 0.20 | 3 | 2 | 4 |
| Cost | 0.15 | 4 | 3 | 3 |
| Safety and waterproofing | 0.10 | 4 | 3 | 5 |
| Weighted total | 1.00 | 3.80 | 3.10 | 3.40 |
Check for the fish-tail column: 0.30×4 + 0.25×4 + 0.20×3 + 0.15×4 + 0.10×4 = 1.20 + 1.00 + 0.60 + 0.60 + 0.40 = 3.80. The paddler scores best on control and safety, but a single-hinge paddle gives less room to test a bio-inspired feature, so the fish-tail swimmer is chosen on the weights that were agreed in advance.
Need help turning your own concept sheets into a defensible selection matrix and design report? Share your brief and we will talk you through the structure.
3. Design requirements
| ID | Requirement (illustrative) | How it is verified |
|---|---|---|
| R1 | Mean forward speed of at least 10 cm/s over a 2 m straight run | Video timing against floor markers |
| R2 | Heading error of no more than 10° at the end of the run | Controller log and video |
| R3 | Hull stays dry for a 15-minute immersion | Visual check after each session |
| R4 | Total mass of no more than 0.9 kg and neutral or slightly positive buoyancy | Scale and tank check |
| R5 | Battery runtime of at least 20 minutes at the test duty cycle | Timed bench run |
4. Analysis: servo torque sizing
The first analysis sizes the tail servo. Assume an estimated peak hydrodynamic drag force on the fin of 0.35 N acting at a lever arm of 0.08 m from the hinge. The peak torque is 0.35 × 0.08 = 0.028 N·m. With a safety factor of 3, the design torque is 0.084 N·m, which is about 0.86 kgf·cm (0.084 ÷ 0.0980665). A waterproof servo rated at 2.0 kgf·cm at the supply voltage gives a margin of about 2.3 times, so it is acceptable. The drag value is an assumption and must be replaced by your own estimate or measurement.
A second analysis, not shown here, would be a buoyancy check against requirement R4: displaced volume multiplied by water density and gravity must match the weight within the ballast adjustment range. State the fluid density you assume.
5. CAD, fabricated parts and tolerances
Two student-designed functional parts are made by approved processes: a 3D-printed servo bracket and hull spine, and a laser-cut compliant fin in 1 mm polypropylene sheet. The drawings need at least three views, an exploded or assembly reference, and at least eight critical dimensions with tolerances. Examples are the 0.08 m hinge-to-fin lever arm (±0.5 mm), the servo bracket bolt pattern (±0.2 mm) and the fin thickness (1.0 mm ±0.1 mm). Tolerances should be tied to function: the bolt pattern matters because misalignment loads the servo shaft, whereas the hull length does not need the same precision.
6. Closed-loop control
The controller uses an inertial measurement unit as the sensor, a microcontroller as the controller and the tail servo as the actuator. The measured quantity is heading, the reference is the desired heading, and the controller adjusts the tail’s centre-line offset to steer while the beat frequency is held constant. A proportional-integral-derivative law is a sensible start for this plant (Åström & Murray, 2021).
| Item | Value (illustrative) | Note |
|---|---|---|
| Control loop rate | 50 Hz | Fast compared with the tail beat |
| Beat frequency | 1.5 Hz | Held constant for all trials |
| Tail amplitude | ±25° | Held constant for all trials |
| Initial gains | Kp = 0.8, Ki = 0.05, Kd = 0.2 | Tuned in the tank, then fixed |
| Logged variables | Time, heading command, heading measured, tail offset | Needed for the performance plot |
The performance variable is the heading error against time. A good plot shows the command, the measured heading and the error on one time axis, so the marker can see settling time and overshoot. Freeze the gains before the experiment, otherwise the controller becomes a confounder.
7. Experiment: compliant fin against rigid fin
The protocol uses the same hull, servo, amplitude, beat frequency, tank, water temperature and start position in both conditions. Only the fin changes. Each condition runs for 10 trials in a randomised order, and each run measures the mean forward speed over a 2 m straight section. The values below are invented for demonstration.
| Condition | Trials | Mean speed (cm/s) | Standard deviation (cm/s) | Standard error (cm/s) |
|---|---|---|---|---|
| Compliant fin (feature ON) | 10 | 12.17 | 0.41 | 0.13 |
| Rigid fin (baseline) | 10 | 10.16 | 0.36 | 0.11 |
| Difference | 2.01 | 0.17 |
The standard error of each mean is the standard deviation divided by the square root of 10, and the standard error of the difference is the square root of the sum of the squared standard errors, about 0.17 cm/s (Taylor, 1997). The compliant fin is 2.01 cm/s faster, which is 19.8% above the baseline (2.01 ÷ 10.16) and about 11.8 standard errors of the difference. In this invented data set, the difference is therefore large compared with the trial-to-trial spread.
Confounders still need discussion: battery voltage sag during the session, water movement left by the previous trial, small differences in the start angle, and timing error from video frame rate. Mitigation includes a rest period between runs, fresh battery checks, a fixed start jig and frame-by-frame timing. A good conclusion states what the data support and what they do not. For example, the test shows a speed gain at one amplitude and frequency, not an efficiency gain across all conditions.
Want your own test protocol, uncertainty working and technical defence checked before submission? Send us your brief and we will walk through it with you.
For another worked engineering extract on this site, see the B50TE heat exchanger design reference extract. If your write-up needs structure, our report writing help page explains how we support engineering reports.
References
- Åström, K. J., & Murray, R. M. (2021). Feedback systems: An introduction for scientists and engineers (2nd ed.). Princeton University Press.
- Ijspeert, A. J. (2014). Biorobotics: Using robots to emulate and investigate agile locomotion. Science, 346(6206), 196–203. https://doi.org/10.1126/science.1254486
- Taylor, J. R. (1997). An introduction to error analysis: The study of uncertainties in physical measurements (2nd ed.). University Science Books.
- Vincent, J. F. V., Bogatyreva, O. A., Bogatyrev, N. R., Bowyer, A., & Pahl, A.-K. (2006). Biomimetics: Its practice and theory. Journal of the Royal Society Interface, 3(9), 471–482. https://doi.org/10.1098/rsif.2006.0127
This model answer was prepared by our editorial team in response to a real assignment brief we received. It shows one way to approach the task and is meant as a study guide, not for submission.
This page is public, so Turnitin has already seen it
Read it for structure, sourcing and depth. Submitting any part of it is caught immediately. Your own brief will differ from this one, and so should your answer.