Scientific Whimsy

jake photo (1)“There are two layers: I’m doing something that is intellectually challenging and interesting that no one has ever done before, and also I get to hang out with some really cool people.”

These are the two driving factors that guide Jake Hillard’s research. For Jake, research is about learning outside of the classroom. And unlike many young students, Jake’s attraction to research didn’t stem from a single field of study, but from the professors conducting the research. The driving force for his beginning research was his conversations with a professor at the University of Utah during Jake’s time in high school. “The best thing in the world is when you can get into a room with someone at the end of the day having talked about ten different projects that are interesting and fun.” Since his interest does not lie in a single topic, Jake has undertaken multiple projects both inside and outside the lab. Outside of the lab, his interests have inspired small projects ranging from unboiling an egg to building a one-wheel self-balancing skateboard using a DIY guide he found online.

Jake’s biggest project took place during his Freshman year at Stanford when he joined the Stanford Space Initiative’s (SSI) and Los Alamos National Lab joint project FINDER (Field Integrated Network Driven Entity Recognition), a project that focuses on creating a network that facilitates cooperation between groups of satellites. Jake used Distributed Sensor Networks, networks of sensor nodes that have equal control over the flow of information between them, to collect and exchange large amounts of data. His goal was to track human movement through a field of nodes. Using Raspberry Pi’s cheap, credit card-sized computers as nodes, he wrote software that “gave them a way to see the world, a way to see friends, and a way with which to exchange information.” Two nodes can share information over a distance of 20 yards, so five nodes chained together can “leapfrog the information” and track movement over 100 yards. Jake recalls that while testing these nodes, “it was really exciting when I’d walk through the field, and each node would notice as I passed within range and indicate that it saw me.”

The applications of Distributed Sensor networks extend to all types of flow analysis involving large amounts of data. They were originally used on a large scale by the Department of Defense to locate the sources of radio transmissions through reverse triangulation. What sets FINDER apart is that it can be created using off the shelf materials and on a much smaller scale. It gives people more functionality and promotes learning by allowing them to build a do-it-yourself distributed sensor network.

Now an Electrical Engineering major in his sophomore year, Jake continues to promote education through his work in the Kovacs Lab on frequency layering in complex systems. He studies systems through oscillators:mechanical or electrical devices that operate through repeated fluctuations in energy between two states (also known as oscillation). For a simple example, think of the pendulum in an old clock. Electric oscillators vibrate when an electric current is applied to them. The frequency of this vibration depends on the material used to make the oscillator and on how the oscillator is made. Normally, one would expect two identical oscillators to have completely identical frequencies. However, oscillators are only accurate within a few parts per million. This may seem to be a small inaccuracy, but the effect is compounded when multiple oscillators are used in a system. These discrepancies can compound enough to render a system ineffective, especially when working with analog control systems, systems that convert information into signals of varying amplitudes (eg. clocks with minute and hour hands) as opposed to digital systems that use binary systems, or on/off switches, to store information (e.g. digital clocks). It is because of these discrepancies that electronics like televisions and computers have moved away from analog circuits to digital circuits that are easier to implement and have reduced error.

In order to describe frequency error, Jake has built a circuit board containing several oscillators. It demonstrates how the addition or multiplication of “identical” frequencies results in layering. When adding frequencies, this layering produces beat frequencies (due to small differences in frequency), and when multiplying frequencies, it produces harmonics. The main role of the circuit board is that of an education tool to help people understand the basic concepts of embedded systems and why frequency error is an important “source of failure.” Research as an educational tool is important to Jake. He believes that, “if you just learn from school, a lot of people’s knowledge is fragile.” A concept may make sense in a given example that is taught in class, but once that example is slightly altered, all perceived understanding dissolves. To Jake, research, is “ a way to apply ideas and explore the world.”


by Ankit Baghel