Research interests:
I am dealing with agent-oriented programing - a kind of programming (suitable for application like
control systems of single mobile robot or modeling of "mind" of living creatures) where software
modules and relations are inspired by multi-agent systems (which are typically
distributed as a team playing robosoccer). We concern this kind of programming as a platform for
building of artificial complex and intelligent systems as well as for modeling of living creatures
(these models are similar to socienty of mind by M. Minsky).
We concentrate on (purely) reactive agents communicating indirectly through (LINDA-like) space
(so called strigmergic communication). For development we have applied
subsumption principle (R.Brooks). This means we have collected in
own platform - the so called Agent-Space
architecture (slovak version, version 2
). We have used this platform for
development of several industrial systems (like pedestrian recognition
- attention: it is not so perfect as it seems from one video!)
and for several models of living systems containing their visualisation in 2D or 3D
(like mating behavior of digger wasp). We are using
it also for control of mobile robots
Besides to further application of Agent-Space architecture, I am partially dealing with XML languages and transformations useful for building real-time system applications.
Since control architectures like Agent-Space are significant only when robot sensors are rich enough, I am dealing also with Computer vision, e.g. OCR of tire (video) and deep learning.
Finally it is much easier to work with simulated robot that with real hardware, thus I am dealing with various simulators and namely with their combination with real environment, like iCubSim (video) or game engines as Unity3D (video)
Hobby: mobile robots - embodied and simulated.
Reimplemented TUTBOT (video),
reimplemented ALLEN in VRML (video),
Jori robot,
Kamko robot,
PingPong - Boebot (video) get sources
Robotour walkway recognition
PingPong II - Boebot (video)
See my GitHub for open source codes.
I read any book about biology or evolution which is available to me and sometimes I model it by multiagent systems or genetic algorithms.