Cats have very expressive ears. The process that they use to control their ears is complex and not fully understood, but they do appear to point their ears toward the sources of interesting noises. With sensors and simple data processing, we can replicate some of this behavior.
Up to eight small directional microphones, placed around the head and pointing in different directions, capture ambient sound. It may be possible to place them under the fur, making them completely invisible.
Each microphone sends an analog signal to two operational amplifiers that preamplify the signal and then filter it to get the amplitude. A microcontroller reads the amplitude from each microphone and uses some kind of algorithm to detect sudden changes in the ambient noise. The software on the microcontroller may command the ears to point towards sound sources.
Our prototype board will demonstrate ear control with one microphone. It features a switching voltage regulator to convert the 12 volt input to 6 volts to power servo motors. The 6 volt supply feeds two independent linear regulators that provide separate 3.3 volt supplies for the analog and digital circuitry. The board can control up to four standard servo motors.
With additional sensors, we can implement other behavior that may be interesting. The Hillcrest Labs BNO080/BNO085 IMU sensor includes an accelerometer, gyroscope, and magnetometer, as well as a microcontroller that processes the sensor data into useful information. It can judge if the sensor is completely still, being held but not moving, or moving. With this information, the ears could relax when the wearer appears to be asleep, and perk up when they start moving.