The computing industry is approaching a strategic point, where a number of changes and limits are coming together at the same time, such as the eventual end of the exponential growth of chip density, and the lack of high availability, fault-tolerance, and maintainability of information technology and computing systems. In the next 10 to 20 years, we will see a major slowdown in the rate of increase in transistor density on the chip due to fundamental physical limitations. The end of the era of exponential growth in transistors (basically Moore’s Law) will force a major restructuring of the semiconductor industry and a complete rethinking of ways of designing high-performance computing systems. This is actually very significant considering that up to now the major observed performance improvements of a single-chip building block have come from increased transistor density (i.e., parallelism and the use of parallel processing techniques) and little from clock rate improvements.
Another change is dictated by the number and complexity of today’s computers and devices. Computer systems have become so large, complex, and fast and have assumed some many important tasks that when things go wrong (and things will go wrong), it is often very hard to implement fixes fast enough to avoid mission-critical problems. The growing torrent of data and the increasing number of connected computers and devices generated by the information technology explosion have added to the problem. This change has spun a revolution in computer design where the emphasis is on self-healing and self-monitoring systems that can detect problems and continue to operate by fixing or by bypassing malfunctions without human intervention. This new paradigm is called autonomous computing by some and recovery-oriented computing by others. Current computer systems are very bad at this since they are based on a technology and computational models, which are fault-intolerant, require high precision, extensive global communication, and are globally synchronized. Researchers are actively exploring possible alternative technologies.
Computing systems inspired by biological systems, Biologically-Based Computing (BBC) are emerging as one possible alternative for tackling the fundamental problems mentioned above. Here we define BBC as the use of biology or biological processes as metaphor in developing new Information Technology-oriented and computing paradigms where information is processed, stored, and transported using natural biological systems. This includes the development of architectures, algorithms, as well as hardware inspired by biological systems. This is in contrast to biocomputing or bioinformatics where computer science and engineering concepts are used to further explore biology.
The complex interactions between living cells and their architecture are emerging as important factors in driving the behavior of complex biological systems. Of particular relevance is the problem of emerging behavior of multicellular structures in which cell-level functions (intracellular pathways) integrate with the architecture through cell-to-cell interactions. Multicellular signaling inherently combines dynamical and structural complexity. More specifically, calcium signaling resulting from the dynamics of endoplasmic reticulae and cell-to-cell communication via gap junctions play a critical role in the behavior of structures composed of endothelial cells.
Our individual laboratories have made recent discoveries and advances in:
(a) The shaping of calcium wave signaling in multi-cellular networks of endothelial cells by the connection topologies of the networks leading to functions such as signal spectral filtering or signal propagation/arrest or signal gating,
(b) The experimental fabrication and engineering of complex networks of endothelial cells and the control and actuation/readout of calcium waves and
(c) The study and design of fault tolerant cellular communication systems.
The objectives of this exploratory effort are three fold:
- To synergistically integrate the fields of multicellular signaling, engineered multicellular topologies, and fault tolerant cellular communication systems to develop novel theoretical and experimental methods to advance understanding and control of calcium signaling in engineered networks of endothelial cells, and
- To shed light on the poorly understood phenomenon of cross-level interactions in complex and dynamic multicellular structures and
- To extend this understanding and the control of multi-cellular signaling to initiate a paradigm shift in the design of fault tolerant cellular communication systems based on an all-biological platform.
Financial support is currently provided by the McDonnell Foundation.
Exploration Team Participants:
P. Deymier (MSE)
J. B. Hoying (Cardiovascular Innovation Center, U. of Louisville)
A. Louri (ECE)
S. Ramasubramanian (ECE)
B. Vasic (ECE)
P.K. Wong (AME)