Engineering Design
Optimization
In
many instance of Engineering design optimization, the simulation of
the problem is often computationally expensive, hence limiting the
scope of design exploration. At Optomatica, we have developed highly
efficient optimization algorithms that work in tandem with meta-models
to learn about the design landscape. This results in a significantly
improved design on a limited computational budget.
Almost all optimizers work in a black box
fashion, where the designer is left with a set of optimal design
parameters at the end of the optimization run. The designer learns
very little about what gave rise to this optimal design and the
interplay this goes on amongst the design variables. At Optomatica we
have developed algorithms and visualization tools that allow for
a better understanding of those interactions and offers a deeper
insight about the problem at hand. Hence, enhancing the creative
design experience and helping the designer cognize the
experience the optimizer.
e-Procurement
The essence of
e-Procurement is optimally matching a buyer's preferences to a
supplier's capability. The richer the language of expressing
preferences and capabilities the more complex is the high dimensional
space where matching occurs, and the harder and slower the
optimization becomes.
At Optomatica we have a
great deal of experience building systems that are expressive and yet
at the same time formulated in such way that greatly reduced their
optimization times.
Network
optimization
Networks permeate all
segments of the economic web, from physical networks transporting gas
and oil, to electric grids, to information and social networks.
As networks come to been as
carriers of complex economic transactions rather than simple
flow amounts the optimization become significantly harder.
Unlike flows, transactions
can interact in highly non-linear ways. A transaction may "bump-out"
another one based on higher contractual priority, or it can cause
other transactions to be reduced using a certain business rule.
Traditional network optimizers are not suited to handling none flow
entities.
At Optomatica we
have developed highly granular network optimization algorithms that
handle such complex network efficiently and allows for a greater deal
of transparency in understanding the effects of various bottlenecks on
the allocation of transactions on the networks.
Organizational design
The power of simulation
techniques combined with ever faster computers has made possible the
modeling of complex organizational interactions. Many problems that
were though intractable can now be ready simulated on high performance
grid computers.
Using agent based
models we can now simulate how different departments within an
organization work together. We can ask questions like: What is the
effect of changing a certain interdepartmental rules on sales
performance? or What is the operational risk of a new business
expansion strategy?
By being able to play many
"what-if" scenarios and applying our optimization expertise to a given
business problem, we can assist businesses in evaluating the
veracity of their decisions and the risks involved.
