Bio-Inspired Track

Connect Digital to Biological

For Biologists and Bio-Inspired Computing Researchers

Ants have been optimizing for 140 million years.

We’ve been coding for 4 days.

What can they teach each other?

The Bridge

What Real Ants Do

• Deposit multiple pheromone types (10-20 chemicals)
• Exhibit tandem running (one ant leads another)
• Use negative pheromones (mark bad paths)
• Adjust behavior based on colony age/size
• Respond to rate of interaction, not pheromone alone

What Our Digital Ants Do

• Deposit one pheromone type
• No direct interaction
• Only positive reinforcement
• Fixed behavior parameters
• Pure pheromone response

There’s a gap. You can close it.

Research Directions

1. Validate the Model

Question: How accurately does our digital stigmergy model real ant behavior?

Approach:

• Compare our decay rates to biological pheromone half-lives
• Compare trail formation dynamics to laboratory observations
• Compare caste ratios to field studies

Deliverable: Validation report with specific recommendations

2. Improve the Model

Question: What biological features should we add?

Ideas:

• Multiple pheromone types
• Rate-of-interaction sensing (Gordon’s key insight)
• Colony age effects
• Environmental responsiveness
• Recruitment behaviors

Deliverable: Specification for biological improvements

3. Make Predictions

Question: What does our model predict that could be tested in the lab?

Approach:

• Run simulations with specific parameters
• Derive testable predictions
• Design laboratory experiments

Deliverable: Experimental design for biological testing

4. Comparative Analysis

Question: What’s the same and what’s different between digital and biological?

Approach:

• Side-by-side comparison of behaviors
• Identify convergences (same solutions)
• Identify divergences (different solutions)

Deliverable: Comparative analysis paper

Biological References

Essential Reading

• Gordon, D. M. (1999). Ants at Work
• Gordon, D. M. (2010). Ant Encounters
• Hölldobler, B. & Wilson, E. O. (1990). The Ants

Key Papers

• Gordon et al. on task allocation via rate of interaction
• Czaczkes on trail pheromone dynamics
• Detrain & Deneubourg on collective decision-making

Lab Connections

If your institution has an ant lab, we want to talk.

What We Provide

Data

• Complete pheromone dynamics
• Agent behavior logs
• Comparison templates

Simulation

• Adjustable parameters
• Custom pheromone types (if you build them)
• Controlled experiments

Collaboration

• Access to CS team for implementation
• Access to Math team for analysis
• Publication opportunities

Challenges

Challenge 1: Biological Accuracy Audit

Assess how accurate our model is.

Deliverables:

• Point-by-point comparison to biological literature
• Accuracy score on key behaviors
• Prioritized improvement list

Prize: $500 + Lab collaboration opportunity

Challenge 2: Multi-Pheromone Design

Design a multi-pheromone system based on biological evidence.

Deliverables:

• Specification for 3+ pheromone types
• Decay rates from biological literature
• Interaction rules between types
• Expected behaviors

Prize: $1,000 + Implementation by CS team

Challenge 3: Testable Predictions

Generate predictions our model makes that could be tested in a laboratory.

Deliverables:

• 5+ testable predictions
• Experimental designs
• Expected results if model is correct
• Expected results if model is wrong

Prize: $1,000 + Funding support for lab experiments

Challenge 4: Biological Improvement Implementation

Add a biological feature to the system.

Deliverables:

• Working code for new feature
• Documentation
• Comparison of behavior before/after

Prize: $1,500 + Co-authorship

Why Biology Matters

Everyone else is building from first principles.

You know what works. 140 million years of evolution already solved these problems.

• Pheromone decay rates? Ants figured it out.
• Exploration-exploitation balance? Ants figured it out.
• Scaling from small to large colonies? Ants figured it out.

Your job: bring that knowledge to the digital realm.

Judging Criteria

Team Composition

Required:

• At least one biologist (entomology, ecology, behavior)
• At least one non-biologist (CS, Math, or other)

Ideal:

• Entomologist (ant specialist)
• Behavioral ecologist
• CS developer (for implementation)
• Data scientist (for comparison)

The Big Opportunity

If you help bridge digital and biological stigmergy:

1. Novel Research Direction — Bio-digital comparative stigmergy
2. Publication Venue — Insectes Sociaux, Behavioral Ecology, PNAS
3. Lab Collaboration — Run experiments with our predictions
4. Career Distinction — Pioneer in a new field

“Ants have much to teach us about building complex systems.”

— E. O. Wilson

They’ve had 140 million years.

We need your expertise to catch up.

[REGISTER FOR BIO-INSPIRED TRACK]