Biomimetic Self-Organization and Self-Healing - Engineered Biomimicry - page 342

Holten-Andersen and colleagues are pursu-
ing ways of synthetically creating polymer net-
works with ligand-metal bonds in order to
achieve self-healing properties [48] . They have
developed a strategy for introducing bis- and
tris-catechol-Fe 3 + cross-links in polymeric net-
works and have demonstrated that they can pro-
duce a polymer network with high elastic
moduli and self-healing properties—much like
the mussel thread cuticle.
The formation of catechol-Fe 3 + complexes
requires an alkaline environment, i.e., pH higher
than 7, but Fe 3 + ion solubility is very low under
such conditions. This challenge is overcome by
prebinding Fe 3 + in mono-dopa- Fe 3 + complexes
by mfp-1 at pH≤ 5. Release into seawater (pH
8) causes the material to spontaneously cross-
link. The resulting substance has the desired
properties, but it is a model gel that requires
further research and development before becom-
ing useful for practical purposes.

They feed on microorganisms inhabitating dead
plant material and reproduce with spores.
Researchers from the Hokaido University in
Japan have found inspiration for the simulation
of traffic networks in the slime mold Physarum
polycephalum [49, 50] . The researchers have devel-
oped a biologically inspired mathematical model
that captures the core mechanisms in the slime
mold's network design [51, 52] .
Conventional network planning requires cen-
tralized control and globally available informa-
tion. In contrast, the slime mold is self-organized
and bases its optimization only on locally avail-
able information. The researchers compared the
network performance in the Tokyo rail network
with a network created by the slime mold. The
Tokyo rail network was mimicked by placing
oat flakes at positions matching the rail stations
and by restricting the growth area using light
( Figure 13.15 ). The slime mold avoids light and
will therefore not grow into exposed areas. The
remarkable results show that the slime mold has
a network performance that is comparable to the
one of the railway network.
Network performance can be measured by
looking at the total cost (total length of connec-
tions), the transport efficiency (average mini-
mum distance between nodes), and resilience,
which is the fault tolerance to accidental discon-
nection. A relative number for each of these
three measures can be found by comparing with
a minimal spanning tree , i.e., the most cost-effec-
tive network that uses as few connections as
possible to link all nodes. A spanning tree is a
graph where all nodes are connected, either
directly or through other nodes. The relative
cost of the slime mold network is about 175%
and the railway network is about 180% of the
minimum spanning tree. The transport effi-
ciency is comparable for the two, since the mini-
mum distance between nodes is only about 85%
of the minimum spanning tree. The resilience is
best for the railway network since only 4% of
faults in the network would lead to isolation of
any part of the network. The same number for

13.8 ADAPTIVE GROWTH

Earlier in this chapter, we discussed how the
behavior of individuals following simple rules
gives rise to emergent optimal properties such as
collective behavioral decisions on optimal forag-
ing and nest-site selection. Similar outcomes are
found if we look at growth in uni- and simple
multicellular organisms, where local growth
rules can result in optimal overall growth pat-
terns without the necessity for gathering any
global information.
Slime molds are eukaryotic (i.e., a cell contain-
ing complex structures enclosed in membranes,
such as the cell nucleus and the mitochondria)
unicellular organisms that were previously
thought to belong to the set of fungi but are now
classified under Protista along with algae. The
slime mold begin life as individual amoebas,
which then mate and fuse together to form
large colonies that usually are up to several cen-
timeters in length but can grow up to a meter.