EvoLib™

EvoLib™ is an artificial life form based on evolutionary computation principles. Our goal is to design software that it is approachable to programmers who lack a specialized background in evolutionary computation. It represents a commercialization of artificial life simulation technologies and is capable of solving many different types of problems without requiring re-engineering of the core engine for each problem type.

The primary advantages of this technology include simplicity of implementation; flexibility; continuous adaptive learning capability including the potential for learning in the field; and significant commercialization potential across several industries.  Potential uses include:

Pattern Recognition

• detecting land mines in airborne multispectral sensor data
• exploring seismic survey data for oil and gas exploration
• airline or cargo baggage screening
• cell counting in pharmaceutical assays
• data mining in clinical study and genomics data sets

Entertainment

• NPC (non-player character) control for role-playing simulations
• An artificial life-based massively multi-player online role-playing game (MMORPG) where everything (trees, grass, animals, NPCs, etc.) is "alive" and evolves.  Players could literally domesticate and breed plants and animals, and the environment would evolve in response to changes brought about by players.
• Opponent in strategy games such as Chess, Go, etc.
• Classifying music from waveform data, or picking music that people might like (e.g. a trainable personal music search agent)
• As an artistic end in itself for the rendering of beautiful graphics or interactive art
  
• Toys whose behaviors evolve, that can be “mated” with each other, can learn or be “domesticated,” etc. (even as a novelty item... think of a much smarter “Furby.”)

Financial Services

• Portfolio optimization and automated investments
• Fraud detection from consumer use patterns or market data
• Market prediction
• Detection of forged signatures for check/signature verification
• Anomaly detection in market data for event forecasting
• Credit scoring
• Evaluating potential investments

Telecommunications

• intrusion detection for network security

Defense

• Target recognition of many types.
• To detect land mines in airborne multispectral sensor data
• Airline baggage screening
• Annotation of satellite imagery for certain features (roads, silos, etc.)
• Control systems for autonomous vehicles depending on the specific need and application
• Playing war games to explore strategies
• Controlling unknown/adaptive agents representing enemies in simulations

Pharmaceuticals

• Visual cell (and other entity) counting for biological assay
• Annotation of patient scan (PET/CAT/etc.) data and recognition of tumors or other diseases
• Searching for active regions in proteins
• Data mining in clinical study and genomics data sets