"Smart" is the new "Tiny." That is both the title of a recent talk given by Prof. Koser, and our overarching motto that guides our research. We at the KoserLab are fascinated by the world of the small, and are driven to study micro and nanoscale natural phenomena in an attempt to transform them into smart engineered systems.

We are an interdisciplinary team, with current and previous members formally trained in the fields of physics, chemistry, and biomedicine, as well as in the disciplines of electrical and mechanical engineering. Prof. Koser himself has both physics and engineering degrees. Nevertheless, our home department is Electrical Engineering, and our eventual focus is the design and creation of tangible systems that may one day help improve lives.

Often, a question comes up in conversation: do we see ourselves as scientists or engineers? Well, to paraphrase a much beloved mentor to Prof. Koser: We have been trained as engineers, and are continually educated as scientists.

Here, you will find a few examples of our ongoing research projects. Please feel free to contact us if you need more information or have questions.


Fig. 1: Our prototype ferro-microfluidic device for cellular manipulation and sorting.

Ferrofluids are colloidal mixtures of magnetic nanoparticles that are covered by a surfactant, and suspended in a compatible liquid medium (see our section on What is a ferrofluid? for more details). When appropriately stabilized, this interesting class of materials responds to magnetic fields while retaining liquid properties. Currently, most conventional applications for ferrofluids involve stationary fields from permanent magnets to hold ferrofluids in place, for such uses as liquid seals and bearings in rotating machinery (see our discussion on Why do we care about ferrofluids?). New research focuses on biomedical applications of ferrofluids, including in targeted drug delivery, immunoassays and contrast agents for magnetic resonance imaging.

Under moderate fields, stable ferrofluids behave as uniform magnetic liquids. This means they respond to externally applied fields by being magnetized. Hence, just as a piece of magnet can exert a force on a piece of iron, external magnetic fields can act on and manipulate ferrofluids. We have discovered that this means ferrofluids can be pumped via magnetic fields without any mechanically moving components. Moreover, anything nonmagnetic, placed inside a ferrofluid, will also feel the effect of these magnetic forces. We are using this phenomenon to manipulate microparticles and live cells directly inside bio-compatible ferrofluids that we have engineered. Below is a list of ferrofluid-related research that we conduct in our laboratory; click on each title for more information.
  • Ferrohydrodynamic pumping
  • Accelerated mass transport via ferrofluids
  • Cellular manipulation, sorting and separation in biocompatible ferrofluids
  • Rapid, label-free pathogen detection in ferrofluids

Smart Sand: Energy Harvesting Wireless Sensor Platform

Fig. 2: An unpackaged Smart Sand device.

We have been developing a silicon-based platform that harvests vibration energy from the environment for a variety of highly integrated, self-powered wireless sensors. It is a millimeter-scale device that utilizes piezoelectricity to convert ambient vibrations into electricity. The device itself provides enough silicon area to integrate all the electronics for power management, sensing and communication, resulting in a very compact sensor node design. Unlike conventional vibration-to-electric conversion system s, these piezoelectric-based micro-scale devices do not require any design customization, mechanical trimming or resonance tuning. Utilizing nonlinear dynamics enables optimal voltage generation from a much larger surface area over an actuator; it also allows effective energy harvesting at low frequencies using micro-scale devices (Fig. 3). Correspondingly, these vibration energy harvesters – named as “Smart Sand” for their eventual small v olume on the order of 1 mm3 – offer a power density improvement of over two orders of magnitude, compared to the state-of-the-art in other vibration-based energy harvester devices. Energy can be efficiently collected from vibrations between several tens of Hz to several kHz, without any tuning. This a spect of the harvester makes it an attractive candidate for a diverse set of real-world applications, including wireless and battery-less sensors for continuous patient supervision outside hospitals, surveillance and security monitoring, border security and structural health monitoring of buildings.

Fig. 3: Unlike linear resonant harvesters that have narrow bandwidths around a sharp resonance (a), our design works off-resonance, resulting in a wide bandwidth (b). Moreover, as opposed to bending energy constrained to the beam base in linear resonant harvesters (c), Smart Sand platform utilizes thin tethers get stretched with large proof mass deflections, allowing virtually the entire beam length to contribute to harvesting.

At Koser Lab, we are not only developing the energy harvesting platform, but also designing and building other related system components (such as various sensor and communication blocks) for eventual integration with the Smart Sand platform. Already, we have demonstrated a low-power temperature sensor that works effectively with varying supply voltages – a condition that is regularly encountered with systems that harvest their energy from stochastic sources, such as ambient vibrations. We are also building power circuitry and switching components that activate the sensor once the harvested energy is sufficient to make a measurement and to transmit it. Already, we have demonstrated a simple electronic switch that turns on the rest of the sensor node only when the collected voltage is large enough to ensure correct operation of the circuit components.

Initial prototypes of Smart Sand devices are currently being fabricated. Our first priority is to characterize these devices and demonstrate the performance enhancements that arise from their nonlinear behavior. We are also building MEMS power switches as alternatives to electronic versions. The main advantage of a MEMS power switch is the lack of leakage currents when the switch is off, saving virtually all the precious stored energy until enough of it is accumulated to power the rest of the circuitry. Accordingly, a MEMS switch will enable sensor functionality in even gentle vibrations. We are also planning to integrate non-volatile on-chip memory with the Smart Sand platform in the next two years. The ability to record measurements for future read-outs will make these sensors practical for a wide range of applications where history logging capabilities are important, such as in long-term patient monitoring. One long term goal of our research on energy harvesting wireless sensors is to eventually create a hybrid network involving both larger, battery-powered sensor nodes with enough power for long range transmission and Smart Sand devices with battery-less operation capability. In the context of a security or surveillance application, the battery-powered nodes could be operated under a deep sleep mode until energy harvesting devices detect activity and send a wake-up signal. This hybrid system offers operational advantages over monolithic networks of either type of sensor nodes, such as highly extended deployment life for the larger nodes and better overall area coverage.

Bacterial Hydrodynamics

Fig. 4: E. coli cell bodies (without flagella) drifting with the flow undergo periodic  orbits that resemble that of a kayak paddle. From  these orbits, the hydrodynamic interactions with the fluid and the underlying surface may be deduced.

Water -- the essential ingredient for life -- is all around us on this planet. Most fish and mammals (including us) that swim in water do so by relying on inertial effects -- such as gliding between fin strokes. At a scale almost 1 million times smaller than us, however, microorganisms have to contend with viscous effects, instead. To them, the same water feels like molasses -- a thick, viscous medium in which the nonlinear inertial effects virtually disappear. There is a great deal to learn about the fundamentals of hydrodynamics under these simplified conditions from such tiny creatures as motile bacteria. We are interested in understanding how they propel themselves, especially near surfaces where additional drag forces are introduced.

Recent research at our laboratory with microfluidic devices has shown that the presence of flow, particularly near surfaces, has a profound effect on how peritrichously flagellated bacteria (such as Escherichia coli) swim. In particular, we have shown that E. coli in flow near a surface exhibit a steady propensity to swim towards one side (within the relative coordinate system) of that surface. This phenomenon depends solely on the local shear rate on the surface, and leads to cells eventually aligning and swimming upstream preferentially along a sidewall or crevice in a wide range of flow conditions.

Our results indicate that flow-assisted translation and upstream swimming along surfaces should be considered in various models of bacterial transport, such as in pyelonephritis and bacterial migration in wet soil and aquatic environments in general. The implications of these findings are immensely significant. The ability of E. coli and other peritrichously flagellated bacteria to “seek” efficient routes to swim upstream under virtually any flow conditions opens up a whole new research area where “flow-assisted bacterial infection” should be studied.

One possible environment that may be conducive to flow-assisted upstream migration in motile bacteria is the Foley urinary catheter. Urinary track infections caused by Foley catheters have been a serious health hazard since routine catheterizations began decades ago. Advancements in materials, conditions of sterility or patient care practices have not managed to overcome this problem. Today, more than 3 billion dollars are spent annually in the U.S. alone treating secondary infections that are caused by bacteria entering the urinary track of catheterized patients. Some of these infections lead to life threatening complications; others may prove fatal. All eventually lead to bacterial strains that are more resistant to our current antibiotic arsenal.

Based on our research, we believe a sizable portion of catheter-associated infections may be explained through flow-assisted upstream migration of bacteria both intra- and extraluminally. Understanding the hydrodynamics of bacterial pathogenesis may lead to better prevention methods that do not necessitate an escalating chemical (i.e., antibiotic) warfare against progressively resistant bacterial strains.