Custom Electric Vehicles 

Electric Vehicles (EV) are known to have three times more energy efficient than petroleum internal combustion engine (ICE) vehicles. This is true if we do not consider electricity generation, transmission and vehicle battery charging efficiency. ICE vehicles have various vendor-specific technologies regarding fuel efficiency including engines, transmissions and their matching. EV, on the other hand, fuel efficiency is largely dependent on the vehicle weight thanks to simplified drivetrain such as direct drive without a transmission. Nevertheless, true energy efficiency of EV should be justified by the source of electricity. This work introduces a custom, light-weight EV prototype for partial solar powered EV. The primary purpose of custom EV fabrication is to collect various data for multidimensional EV energy and power consumption characterization and modeling. Use of high-fidelity model will be the basis of systematic EV design time and runtime optimization. Furthermore, the custom EV will be a useful testbed for vehicle energy harvesting including dynamically reconfigurable photovoltaic arrays. 


 Power and Thermal Management for Smartphones

 This work introduces a novel sensor-less, event-driven power analysis framework for providing highly accurate and nearly instantaneous estimates of power dissipation in an Android smartphone. The key idea is to collect and correctly record various events of interest within a smartphone as applications are running on the application processor within it. This is in turn done by instrumenting the Android operating system to provide information about power/performance state changes of various smartphone components at the lowest layer of the kernel to avoid time stamping delays and component state observability issues. This technique then enables one to perform fine-grained (in time and space) power metering in the smartphone. Experimental results show significant accuracy improvement compared to previous approaches and good fidelity with respect to actual current measurements. The estimation error of the proposed method is lower by a factor of two than the state-of-the-art method.


Reconfigurable PV

Photovoltaic (PV) energy generation techniques have received significant attention since they utilize the abundance of solar energy and can be easily scaled up. Thanks to the extensive research and development of PV energy generation technologies, various scales of PV energy generation systems have been deployed for many practical applications such as PV power stations, solar-powered vehicles, and solar-powered heating and lighting appliances.


Battery Variation

Cell-to-cell variability of batteries is a well-known problem especially when it comes to assembling large battery packs. Different battery cells exhibit substantial variability among them due to manufacturing tolerances, which should be carefully assessed and managed. By joint research with Politecnico di Torino, we propose a combined cell-to-cell variability model of the capacity and of the internal resistance of a Li-Ion battery that accounts for variability effects in the cell manufacturing process. The proposed model allows us to qualitatively verify some known properties such as the correlation between capacity and internal resistance, and quantitatively assess the amount of variability and its impact on the design of battery packs. 

Battery Aging

Battery-operated portable electronics, from smartphones to notebook computers, are generally sold with a dedicated power supply. The power supply operates the device and also charges the built-in battery. Most users are concerned about the battery aging while the device is operated by the built-in battery. This is the first paper to our knowledge that discovers, analyzes and mitigates the built-in battery aging when the device is operated with the provided power supply. We focus on the fact that in an effort to reduce size and weight, the capacity of the power supply is optimized for the average power demand rather than the maximum power demand. Such a reduced-capacity power supply brings advantages in terms of size, weight and cost but it accelerates the battery aging because the aging progresses even when the device is operated by the power supply, which is different from the expectation of most users. We quantitatively analyze such battery aging with various operating scenarios based on standard benchmark programs. We show that the battery experiences significant aging, i.e., the battery lifetime can be reduced to 23% of its shelf lifetime. Finally, we propose a cost-effective supercapacior hybrid to mitigate such battery aging when the device is operated using the power supply. The simulation results show that 10, 1 and 0.1 mF supercapacitors can reduce the battery aging by 68.6%, 55.1% and 4.6%, respectively.



Non-Volatile RAM

Emerging memory technologies including phase change memory (PCM) have been researched to the limitation of the conventional memory devices. However, most previous literatures have the less addressed  practical issues such as LPDDR2-NVM which is an industry standard interface of NVMs. It has been adopted in most commercial PCM prototypes. The research of our laboratory has been focused on the practical issues of emerging memory technology including LPDDR-2 NVM standard interface, row buffer management, addressing acceleration, and other performance enhancements of the PCM devices. We have implemented a PCM-based embedded prototypes consisting of a LPDDR2-NVM compatible PCM controller, array of PCM SODIMMs and other peripherals. This prototype has been used for verifying the functionality and effectiveness of all our research ideas.

Hybrid Energy Storage Systems

An ideal energy storage is expected to have high energy density, high power density, high cycle efficiency, long cycle life, environmental-friendliness, low cost, and so on. However, today’s energy storage technologies have advantages in some aspects while having limitations in other aspects. Hybrid energy storage systems (HESS) exploit multiple, heterogeneous energy storage technologies in a complementary way to take advantages of the strengths and hide the weaknesses of each technology. The HESS systems practically achieve near-ideal performance with the currently available technologies. Our research focuses on various high-level design optimizations and runtime managements to maximize the benefits of the hybridization.





Storage-less Converter-less MPPT and SmartPatch

Long-term energy storages and voltage level converters have been necessary components in conventional energy harvesting systems even though those components significantly limits the lifetime, cost and form factor of the systems. Our recent research proved that removing those components is possible by directly supplying the harvested solar energy to the computing node.  For proving the functionality and effectiveness of the proposed storage- and converter-less energy harvesting architecture, we have implemented a practical circuit of power management unit (PMU) capable of performing maximum power point tracking (MPPT) of solar energy without requiring any long-term energy storages and voltage converters.