# Smart Bricks Mimic Biological Intelligence to Self-Assemble

Neural network-powered cubes identify and regenerate their collective shapes with 100% accuracy.

- **Source:** [Scientific American Content: Global](https://brifing.si/source/4f066134-e06d-4415-87ef-305e3dc2aba6)
- **Published:** 2026-09-06
- **Topic:** [Data & AI](https://brifing.si/c/data-ai)
- **Original article:** <https://www.scientificamerican.com/article/these-smart-bricks-know-what-object-they-make-up/>
- **On Brifing:** https://brifing.si/s/d08791db-6dc7-4806-a73d-9a946956a995

## Biological Inspiration

Researchers mimicked how cells self-organize without central control to create complex organisms. This inspired 'smart bricks'—cubes with independent neural networks that collectively determine their assembled shape, similar to biological systems. The study, published in Nature Communications, demonstrates 85-100% accuracy in shape identification across seven categories including boats and houses.

## Neural Cellular Automata

Each cube contains a neural network that updates its memory vector based on neighboring cubes, forming a decentralized decision-making system. This approach builds on concepts like John Conway's Game of Life (1970), where simple local rules create complex global patterns. The system requires no top-down instructions, making it resilient to single points of failure.

## Shape Identification

In simulations with nearly 500 shapes, cubes accurately identified their collective form using only local neighbor data—even with 15% of cubes missing. Larger systems with over 18,000 cubes maintained this capability. Houses proved most challenging at 85% accuracy, while simpler shapes like chairs achieved perfect recognition.

## Self-Repair Capability

The system could detect missing cubes and regenerate correct neighbors autonomously. Physical prototypes with centimeter-scale blocks achieved 100% shape recognition when assembled into objects like guitars or tables. This demonstrates potential for real-world applications where components must adapt to damage.

## Edutainment Applications

Lead researcher Sebastian Risi highlights near-term uses in educational toys—like dinosaur kits that 'roar' when correctly assembled. Smart construction bricks could guide users during assembly by indicating where to place the next piece, blending learning with interactive play.

## Modular Robotics Vision

Long-term goals include self-assembling modular robots that require no manuals. Such systems could revolutionize manufacturing by adapting to damage or changing tasks. The research bridges AI and materials science, offering a blueprint for future 'living' machines.

## Technical Breakthrough

The study advances swarm robotics by proving small, simple units can achieve complex collective intelligence. Unlike traditional robotics needing centralized control, these cubes use emergent behavior—a paradigm shift toward more resilient and scalable systems.

## Future Implications

This technology could enable self-healing circuits that repair damage autonomously, akin to biological tissue. Beyond robotics, it may inspire new approaches to distributed computing and adaptive infrastructure, where components collaboratively maintain system integrity without human intervention.

## Sources covering this story

- https://www.scientificamerican.com/article/these-smart-bricks-know-what-object-they-make-up/ — 

---

_Summary written by Brifing from the original reporting. Copyright in the original article belongs to its publisher; read the full piece at the link above._