MITS5509 Intelligent Systems: Anytime vs. Real-Time Search Report

This report analyzes the efficiency of anytime and real-time heuristic search algorithms, focusing on their application in robotics and online planning. The research aims to determine the more effective search algorithm by evaluating their performance across various simulated domains, including an under-actuated double pendulum, a point robot with inertia, and grid pathfinding. The study reveals that the real-time heuristic search approach is more beneficial, offering faster goal achievement compared to the anytime search approach. The report emphasizes the importance of heuristic search in robotics and concludes that real-time searching algorithms are highly effective, although it acknowledges the limitations of incompleteness. The findings highlight real-time search as a promising direction for future research in heuristic search applications within robotics.

ARTIFICIAL
INTELLIGENCE SYSTEM
Anytime versus Real-Time Heuristic Search for On-Line Planning

INTRDUCTION
In the field of robotics the anytime search is one of the
most popular the popular process which is required to
the quickly find something. However, there are several
limitations present which impact the working process of
the AI system.

RESEARCH OBJECTIVE
The aim of this research is mentioned below:
 Analyze the efficiency of any time searching algorithm.
 Analyze the efficiency of real time searching algorithm.
 Identify more effective searching algorithm.

Robot
An automated machine that is formed by using computer programming to
perform any task with human capabilities.
In the field of robotics the there are several searching approaches present which
includes the any time searching approach and real time searching approach.
The anytime searching approach is the most popular searching approach in
robotics so far.

ANYTIME V/S REAL-TIME
HEURISTIC SEARCH
Anytime Heuristic Search- is such searching approach that is very effective in
searching the suboptimal plan then proceed for the better plan of decision
making.
Real-time Heuristic Search- is such searching method that is bounds real time
with action selection time.

METHODS
The performance of real time and any time search approach is measured by
implementing it on Kotlin with numerous simulated domains which includes:
 An under-actuated double pendulum.
 A point robot with inertia.
 A point robot in continuous space.
 Grid path finding in inertia.

RESULT
After the experiment it is identified that real-time heuristics search plan is the
most beneficial searching approach that is high effective is several domains. The
study has also mentioned that bas compared to the anytime searching approach,
real-time search approach is faster as well as effective in achieving the goal.

RESULT
GAT on a slalom graph
consisting double integrators.

RESULT
GAT for grid path finding.

DISCUSSION
In the field of robotics searches the application of heuristic search has been
utilized. And after analyzing this it is noticed that the real time searching
approach is ore effective than the any time searching approach as the goal
achievement time of real time searching approach is better than the any time
searching approach.

CONCLUSION
This research has concluded that for future research on the heuristics research
approach of robotics the application of real time searching algorithm will be
very effective.
However, the primary limitation of this approach is the incompleteness of the
real time searching application.

Bibliography
Cserna, B., Bogochow, M., Chambers, S., Tremblay, M., Katt, S., & Ruml, W. (2016, June). Anytime versus real-time heuristic
search for on-line planning. In Ninth annual symposium on combinatorial search.
Ahmed, M. S., Mohamed, A., Khatib, T., Shareef, H., Homod, R. Z., & Ali, J. A. (2017). Real time optimal schedule controller
for home energy management system using new binary backtracking search algorithm. Energy and Buildings, 138, 215-227.
Khemiri, R., Kibeya, H., Loukil, H., Sayadi, F. E., Atri, M., & Masmoudi, N. (2018). Real-time motion estimation diamond
search algorithm for the new high efficiency video coding on FPGA. Analog Integrated Circuits and Signal Processing, 94(2),
259-276.