Nanoengineering applications potential of the LARCA research group of UPC

Jose L Balcazar, LARCA research group, LSI department, EPSEVG

June 22, 2005



ABSTRACT. It is briefly argued here why the LARCA research group should be 
considered as a potential contributor to the new Nanoengineering
Initiative of UPC.




Nanoengineering could be understood along two somewhat different tracks.
On the one hand, a natural interpretation is engineering under absolute
metrics in the order of the nanometer. LARCA has no potential along this
track. However, a alternative natural interpretation is relative: where 
the problematics are not due explicitly to the fact that nanometric scale
components are used, but to the fact that such components are mainly 
useful due to the possibility of being combined in extremely large
quantities.

Consider a device built up of components of nanometric scale, and being 
itself of absolute metrics on the order of centimeters to decimeters 
(only some exceptional objects, such as the Mare Nostrum, live on a scale 
of meters and are simultaneously composed of nanometric scale parts). Many 
of the engineering problems related to such an object regard the fact that 
the ratio between the scale of the object and the scale of the components 
is about a -7 to -8 power of 10, since this is the fact that allows, 
in principle, for the object to be composed of small parts numbering 
a 7 to 8 power of 10 in a linear configuration. Of course this evaluation 
is just a theoretical gross estimate, and more sensible numbers would be 
much less; but this can be easily compensated by the usual plane (nonlinear) 
organization that adds a square operation, and we should not forget that 
the potential of 3D euclidean space suggests even a cubic increase as a 
potential future step. Assuming, for instance, that we get a factor of 100
below the gross limit, the number of nanoparts on an approximately planar 
organization (several but not many planar layers) could easily reach the 
order of a tenth power of 10.

How to engineer a system in which ten thousand million individual components 
(european terminology - american terminology would be ten billion) must 
interact and cooperate to construct a useful engineering product, able to 
hide the tremendous number of interactions taking place inside into a 
humanly usable (thus, reasonably simple) behavior? Here we find the point 
where, for some nanoengineering problems, the absolute metrics sometimes 
can be factored out as irrelevant: in order to design a way of governing 
all this interaction, many problems lie in the presence of ten-to-the-tenth 
power components, whether they are in the absolute nanoscale and composing
a human scale object, or whether they measure themselves one to two meters 
and compose a planetary society.

For this family of problems of 'relative scale' nanoengineering, we find 
immediately ourselves pretty close to Complexity Theory. Just about twenty 
years ago, a hundred megabytes of computer storage were only available to 
some lucky persons working on computer research environments; but a few 
weeks ago, I paid 20 euro for a second-hand device allowing for well over 
that storage on flash memory. I write this text on an old laptop that 
already had over 20 gigabytes of hard drive. This full potential of storage, 
compared to individual humanly sensible semantic units such as these very
words I am using, yields a ratio precisely of the order of ten to the tenth 
power. Similarly, the exploration of solution spaces for naturally occurring 
combinatorics problems such as Constraint Satisfaction (arising in many 
practical applications, ranging from compiler design to operations research 
and logistics) leads to similar numbers: on, say, 50 binary variables, the 
best recent slightly subexponential algorithms have to deal with explorations 
of spaces of size close to 2 to the 25th power, which is the same scale of 
ten to the tenth power. 

Again, in data mining problems, many datasets arising in more and more
applications of computing sciences exhibit this same order of magnitude, as 
ratio to their individual natural units. And some scientific experiments 
gather even considerably larger quantities of data, whose exploration in 
search of scientific facts raises challenging difficulties very similar to 
those raised by the problem of organizing nanoscale components into meterscale
constructs; the proceedings of the annual international conference Discovery 
Science present many more examples of the opportunity of this scientific 
endeavor. 

This writer has contributed papers to that conference in several opportunities 
in the past, and belongs to the Selection Committee for this year; and the 
edition of 2006 of the conference has been already put in charge of the LARCA 
research group to take place in UPC. Members of LARCA are currently attacking 
a problem of analysis of systems made up of extremely small units (namely 
2-state cellular automata) using novel techniques of closed sequences and 
colimit operations of category theory (submitted to the European Conference 
on Complex Systems). Members of LARCA participate in an Integrated Project
of the VI framework of the UE by the name Dynamically Evolving Large-scale 
Information Systems whose baseline is parallel to the phenomena described 
above.

Thus, even if it will not be easy due to the very different origins and 
languages of the researchers, I am convinced of the opportunity of including 
the LARCA research group in the Nanoengineering Initiative of UPC.