BioWall

Webmaster

Overview of the BioWall

In the Logic Systems Laboratory (LSL) of the Swiss Federal Institute of Technology in Lausanne (EPFL), we have been working on bio-inspired hardware for several years. In our research activities, we have covered most of the possible avenues for such inspiration, ranging from phylogenetic systems, inspired by the evolution of biological species, through ontogenetic systems, inspired by the development and growth of multicellular organisms, to epigenetic systems, inspired by the adaptation of individuals to the environment.
Among all these research axes, the main effort in our lab has been concentrated on the ontogenetic axis, through the Embryonics (embryonic electronics) project, which aims at drawing inspiration from the development of multicellular individuals in order to obtain in digital hardware some of the features of biological organisms, and notably growth and fault tolerance.
Our activities have attracted a flattering amount of interest in the most varied and sometimes unexpected milieus. Among the most unexpected was undoubtedly Mrs. Jacqueline Reuge, who decided to fund the construction of a machine to display the principles of Embryonics to the public within a museum (the Villa Reuge) built to honor the memory of her late husband. Her generous support has allowed us to maintain our tradition of always verifying in hardware the concepts developed for our project.
This serendipitous event allowed us to construct a machine that would otherwise have remained a dream. We named this machine BioWall because of its biological inspiration on one side, and because of its size on the other. In fact, the main goal of the machine being as a platform to demonstrate the features of our Embryonics systems to the public through a visual and tactile interaction, the final implementation of the BioWall weighs in at an impressive 5.3mx0.6mx0.5m=3.68m3 (130 cubic feet).

On this machine we implemented, for the first time in actual hardware, an organism endowed with all of the features of an Embryonics machine. The functionality of this organism, the BioWatch, is to count hours, minutes, and seconds, and is used to demonstrate the growth and self-repair capabilities of our systems.
In a sense, the implementation of the BioWatch would by itself be sufficient to justify the effort that has gone into the construction of our BioWall (the realization of Embryonics systems was, after all, the goal of the machine). However, in developing our machine, we quickly realized that the capabilities of such a platform were not limited to a single application. In fact, as the technical description of the machine should reveal, it is an ideal platform to prototype many different kinds of two-dimensional cellular systems, i.e. systems composed of an array of small, locally-connected elements.
The applications that correspond to this description are numerous, and particularly in the domain of bio-inspired systems. For example, cellular automata (CA) are a very common environment in bio-inspired research, from the classic Game of Life of John Conway, through self-replicating loops as first developed by Chris Langton, to Von Neumann's universal constructor , to name but a few (in growing order of complexity). And while the BioWall is ideally suited to the implementation of CAs, it is by no means limited to it. As examples of other possible bio-inspired systems, we will describe an implementation of a particular type of artificial neural networks, developed by Alan Turing, and a two-dimensional realization of Firefly, a machine we designed and built to demonstrate the feasibility of online hardware evolution.
These applications are just a small sample of the capabilities of the BioWall, capabilities that we are still discovering. The cellular structure of the machine make it an ideal platform for the prototyping of bio-inspired systems, which often exploit this kind of structure, very common in nature at all levels. Its size and structure impose a certain number of limitations (e.g., clock speed), but its complete programmability provides an outstanding versatility (the different applications we mentioned should be a sufficient, if incomplete, example) and the visual and interactive component of the system are invaluable tools both for the dissemination of ideas and for the verification of research concepts that are often limited to software simulations.
To conclude, we would like to invite all of you to come and "play" with the machine at one of the events in which it will be in public display or by contacting any member of the staff to arrange a visit to our laboratory. And, on a more "serious" note, we would be extremely interested in putting our machine at the disposal of other research groups, who could be interested in a hardware realization of their ideas and concepts.

Resources