A 50-Year-Old Assumption Overturned
For over 50 years, researchers believed that fish swam in flat diamond formations to conserve energy. This belief was based on 2D modeling and limited experimental data. However, a new study led by Radhika Nagpal, Professor of Robotics at Princeton, and George Lauder, a biologist at Harvard, has uncovered a very different swimming behavior.
Using advanced 3D computer vision software, the research team observed giant danios swimming in a specialized tank over 10 hours. They found that the fish swam in a 3D staggered pattern, which they’ve termed the “ladder formation.”
📊 Key Findings from the Study
- Fish swam in ladder formations 79% of the time.
- Traditional diamond formations were rarely observed.
- Fish swimming behind others generated rearward jets, similar to jet engines.
- Ladder formations allow fish to avoid turbulence while gaining hydrodynamic advantages.
- Unlike diamonds, ladders allow multi-plane staggering, making it easier to maintain formation without exact synchronization.
🔬 How the Research Was Conducted
The team developed a machine vision pipeline to track the 3D positions and movements of fish:
- 6 giant danios were placed in a recirculating flow tank with a mesh chamber to reduce boundary layer effects.
- High-resolution video tracking captured the nose positions of each fish in real-time.
- The software generated accurate 3D coordinates and movement trajectories for each individual.
📽️ Example 2-minute video footage of tracked fish schools confirms the ladder formation pattern.
🤖 Implications for Robotics and AI
This discovery doesn’t just transform our understanding of schooling behavior in fish—it also has important applications in robotics and artificial intelligence. According to lead author Dr. Hungtang Ko, postdoctoral researcher at Princeton:
“We can apply these findings to design underwater robot swarms that mimic the ladder formation for efficient and synchronized movement.”
The Nagpal Lab is already working on fish-inspired robotic systems that could one day help monitor coral reefs, kelp forests, and other delicate marine ecosystems.
📚 Citation and Further Reading
Ko, H., Nagpal, R., Lauder, G., et al. (2025). Beyond planar: fish schools adopt ladder formations in 3D. Scientific Reports. DOI: 10.1038/s41598-025-06150-2
🧠 Why This Matters
This study represents a paradigm shift in marine biology, proving the importance of analyzing animal behavior in three dimensions rather than relying on flat, outdated models. It also highlights the power of computer vision in tracking natural movement, opening doors to bio-inspired engineering and robotic innovation.