Introduction to Biological Sensors

In the study of sensors and sensing technology, we often encounter physical limits to the measurement of certain signals. For example, temperature measurement can be limited by thermodynamics - the presence of temperature fluctuations in all finite objects at finite temperatures. More often, the information which can be derived from a temperature measurement is limited by environmental disturbances - often referred to as environmental noise. For example, the jungle environment naturally produces a broad spectrum of audio background noise, and it is not generally useful to have sensitive hearing capabilities which are continually swamped by background noise. In the case of light sensors, the limitations to use in daytime are usually due to the accuracy of the focusing elements, or perhaps the clarity of the atmosphere (more of a problem these days, it seems). At night, the optical systems of nocturnal animals can be limited by the very few numbers of photons that are available to detect. Human eyes are capable of detecting statistical variations in the intensity of light from stars at night. The "flicker" that you can see in faint stars on clear nights in the mountains (especially in winter!) is due to a combination of atmospheric turbulence and the fact that there are so few photons that statistical effects in the arrival rate are detectable.

The natural evolutionary process in nature has been a remarkably efficient filter for useful sensing capabilities. There are many examples of highly adapted sensing capabilities which offer real performance advantages. And, there are many examples of rather primitive sensing capabilities which can be utilized effectively by the more sophisticated software systems that are also part of most natural systems. Some good examples of high-performance sensing systems include chemical sensing of pheromones in moths, ultrasound sensing in bats (predators) and moths (prey), and vision in predators such as cats and birds. Some examples of low-performance sensors include tilt sensing in humans and tactile sensing on various parts of the anatomy.

To begin to get a feel for sensing technology and capabilities, I want to spend some time in the first lecture experimenting with human sensory capabilities. As we perform these experiments, we want to think about many of the system issues associated with sensing, such as signal conditioning, wiring, time response and hysteresis, packaging, and the role of software/hardware tradeoffs in system operation. We will find that the human sensing systems provide several really interesting examples of ways to convert physical signals into information, and many interesting examples of how to trade off various aspects of system performance and operational cost to achieve the desired result.