Battery Management System

The battery industry is leading the market at a point where not only electric vehicles and energy storage systems but also eco-friendly public policies are combined. In order to diagnose the state of a battery, such as voltage, current, temperature, etc., are received from a cell through cell monitoring, and using this, the remaining capacity and the aging state of the battery are estimated. Our main interest is to implement the properties of BMS with high stability and to design chips and systems that can realize all the functions, including power-related control. BMS is applicable everywhere power needs to be managed, but the most concentrated parts of it are electric vehicles and energy storage systems, and BMS can also be applied to IoT devices.  BMS is a system that includes some kinds of PMIC (e.g. power converter, low dropout, switching capacitor), but is a larger system that includes the below functions. 

Optimized Multi-Storage Control

Since numerous storage devices are managed simultaneously, balancing between modules is essential, and a combination of data transmission and cell status prediction becomes important. Data is transmitted and received in a very hash environment, and it is necessary to design a strategy to balance the state of a cell in consideration of both charging and discharging states.

We build a multi-storage platform that processes quickly from collecting multiple types of energy to transferring energy to multiple loads. Based on the collected information, circuits and systems are configured to be the most efficient and safe. The development of cell state prediction techniques to maximize the reamining useful life (RUL) and techniques to optimize the computational cost is carried out. It is a necessary technology to develop beyond the existing battery managment system (BMS) into the next generation of innovation, and aims to reorganize the entire system.

State Monitoring System

To design a safe system, it is required to be able to verify the safety scale through individual battery monitoring. Various methods, such as electrochemical impedance spectroscopy (EIS), which does not use a sensor are applied to accurately measure a voltage, current, and temperatrue, and a goal is to derive accurate results with a simplified technique through circuit-based lumped battery modeling.

Our lab is performing tasks that allow continuous measurement in real-time with minimal sensors and high stability. The major issues for the state monitoring are the relaxation time of the activated battery, the wide operating range of voltage and current, the noisy envirionment, the mismatch in the factory connection state, and the number of wirings. Before commercialization, we test how safe the designed chips are in different environments. We need to design systems in advance against these accidents in the design process.

IoT-Level Application

As localized digital equipment diversifies in modern society, it becmoes essential to manage batteries in IoT-level devices, such as biomedical devices and smartphone-sized devices, not only in automobiles and energy storage system (ESS). Currently, a battery managment system is being designed for a method of efficiently driving a power source to store energy from a harvester to a battery and transmit various types of information through continuous Blutooth communication. In addition, since power must be transmitted with various types of loads, a circuit that can convert power with a small area is being designed in our lab.

High-Speed Transceivers

In communication systems, transceivers are in charge of transmitting and receiving signals. As data signal is transmitted through medium, such as wireline, optical fibers, and air (wireless), it suffers from distortion caused by various reasons, resulting erroneous data at receiver. For high-speed communication, the impairment is severer and compensation is critical for acceptable signal integrity. For accurate transfer of data, various algorithms and techniques for compensating distortion need to be integrated in transceiver ICs. 

Processing-in-Memory (PIM)

Because of tremendous growth in requirements on data acquisition, processing, and storage bandwidth, the demand of higher speed link that used data transmitting and receiving between IC chips is increasing. On the other hand, in contrast to the increase in the processor speed, the transmission/reception speed between the memory and the processor is slow and a lot of energy is consumed. To eliminate this inefficiency caused by performance gap mentioned above, PIM (processing in memory) architecture is presented.

PIM architecture can reduce the amount of memory access of the processor by adding computing unit inside of memory. If a multi-pim structure is formed using a chip interconnection method such as 3D/chiplet, crosstalk or ISI between channels inevitably occurs, and the quality of I/O signals transmitted through the interconnection deteriorates. We aim to enable high-speed PIM architecture by analyzing and compensating the physical characteristics of signal distortion when using such interconnection method mentioned above.

Electrical Dispersion Compensation (EDC) for Optical Communication

In optical communication, data is carried using light through optical fibers. Data transfer using optical fibers has advantages of higher data throughput, cost effectiveness, and immunity to electromagnetic interference. Therefore, it is widely employed for massive data communication. During the transfer, the light signal experiences chromatic dispersion (CD), which pertains to different optical frequency components of the signal, or polarization mode dispersion (PMD). As data rate increases, the effect of dispersion is so severe that compensation is inevitable. In our lab, we analyze dispersion and develop methods to electrically compensate them. We expect our research to improve energy efficiency and cost-effectiveness of optical communication.

Digital Healthcare

The need of biomedical treatment for various diseases, including Covid-19, is increasing in this worldwide aging society. Instead of heavy health-checking machines in the hospital, we would like to simply monitor our physical condition using our cell phones at home. Our group has interest of small, low energy circuits for biomedical system, as well as real-time monitoring. We study flexible, skin-attachable systems that can sense various conditions in our body such as sweat & blood flow, wound healing state, respiratory rate, etc.

Soft, Flexible Electronics

Major drawbacks of traditional printed circuit board(PCB)-based biomedical devices, is that they cannot overcome the mechanical mismatch between soft human bodies and hard devices. Our group seek to create soft, biocompatible electronics that could measure various body condition. By utilizing low-energy, skin-like flexible electronics allowing close contact with human skin, accurate biological signal measurement is possible for a long time period.

Soft and flexible electronics are envisioned as the future of electronic devices that can be easily deployed in a variety of environments, such as on-body, on-skin and as a biomedical implant. Our group is developing biomedical devices such as sweat & blood flow, wound healing state. We aim to investigate strategies to get physical characteristics of the skin and body via circuit design and BLE communicaiton. An wireless and non-invasive epidermal device for blood flow monitoring and wearable microfluidic device for sweat analysis are currently under study.

Miniaturized Capnography System

We are currently investigating on development of a small devices that monitors end-tidal CO2 concentration. Continuous monitoring of CO2 is carried out by circuit composed of signal amplification and NDIR principles, followed by digital processing. Board development for BLE communication connects signal to portable device, then briefly shows the result to user via Android application.

Wireless Cardiac Monitor

Stent Insertion Surgery is widely used for treating coronary artery spasm. However, it is important to consider the restenosis of coronary artery after surgery. To monitor restenosis, inductor-capacitor based Rx pressure sensor can be integrated with stent. Resonant frequency of Rx sensor can be interrogated from External Tx coil by using wireless power transfer (WPT). To increase readout range considering the distance from heart, Coil geometry and WPT system should be optimized. Also, various integrated circuits can be designed to make portable and robust system.