Welcome to the website for Prof. Andreou Lab at Johns Hopkins University. The Lab is co-directed by Prof. Andreas G. Andreou and Prof. Philippe Pouliquen. The research in our lab comprises of projects with intertwined science, engineering and technology objectives.
From a science perspective our research aims at developing a theory of computation that incorporates constraints imposed by the structures that underly processing, communication and memory. Furthermore we seek to understand how one can do parallel processing effectively and efficiently under the fundamental limitations of delay and energy. While understanding the physical and abstract computational structures of brains is our ultimate endeavor, we are natually deeply involved with the challenges of computing in the era beyond Moore's law. Research in sensory communication is aimed at gaining further understanding of the basic underlying processes in vision and audition as well as the sensory motor integration loops.
From an engineering perspective, biological systems serve as working models of microsystems for sensory information processing and memory as they have been optimized over millions of years of evolution and provide key insights on algorithms, signal representation and architectures. By innnovative use of CMOS state of the technologies, including 3D, mixed signal circuit design techniques and by careful interpreting the organizing principles in neural systems, we are making progress engineering microsystems in the two broad areas of information acquisition/transduction, and knowledge representation/memory and learning.
"50 years from now, a century would have passed from the invention of the first microchip and Moore’s law will be no more. However we will be living in an era where the ‘chip’ – short for the microchip – in its different forms from imagers to ion sequencers to labs-on-chip and cognitive processor units (CogPUs) will provide the underpinnings and the foundation for affordable, state-of-the-art, global personalised medicine and healthcare delivery."Please read more in " Johns Hopkins on a Chip: microsystems and cognitive machines for sustainable, affordable, personalised medicine and healthcare".We are gratefull and thank our sponsors that provide funding for our research: DARPA, NSF, NIH, ONR, AFRL, JHU-APL. If you think what we do is worthwhile and want to support it, please do here or by contacting Rising to the Challenge!
Lab logo: From 1990 article " A Chip You Can Talk To" by Rachel Nowak published in Johns Hopkins Magazine. Rachel ruminates and writes about the future of the Andreou lab research at Hopkins. 30 years later she was right on!
Images on Website Carousel
Sounds of the Heart: Research on physics based modeling and machine learning for heart sounds. Relevant two publications on computational modeling of heart murmurs and the award winning Stethovest, a wearable acoustic sensing array.
Human Centric Technology: Human action recognition without a camera using low dimensional micro-Doppler signatures. Three key publications: acoustic micro-Doppler, model based and deep neural networks algorithms for human action recognition.
3D and Chiplet Based Systems: The early days of 3D CMOS research in the lab involved silicon on sapphire CMOS and VCSEL arrays for optical interconnects, 3D interchip capacitive data/power transfer, and 3D silicon brains using the experimental MIT/LL 3D SOI CMOS technology. More recently we have used the mature Tower/Jazz 3D CMOS available through Tezzaron for computational imaging and 2.5D chiplet SOC design and interposer fabrication.
Extreme Environment Microsystems: Research on science and technology for space based applications, started with the collaboration with Robert Jenkins and later with the friend and colleague, the late Kim Strohbehn of the Applied Physics Laboratory. Early work focused on radiation hard systems for long term space missions while more recent work had emphasis on the science and technology for miniaturized quantum sensing in space, an absolute atomic magnetometer. The device has a noise floor of 15pT sqrtHz; a slightly modified device is sensitive enough for room temperature magnetoencephalography i.e. can be used to measure magnetic fields in brains at room temperature and natural environments.
Brain Science and Technology: Research in our lab is aimed at understanding the fundamental limits in information processing in the brain for example, developing a communication channel model for the blowfly photoreceptor to calculate its information capacity and modeling/computing the energy efficiency of a spiking neural link. Our lab is a member of the Johns Hopkins University Kavli Neuroscience Discovery Institute.
Congratulations to Dr. Jonah Sengupta for successful Ph.D Dissertation defense.Posted by Andreou Lab on Thursday, October 6, 2022