Modular self-reconfiguring robots are composed of repeated modules that can rearrange their connections to form different structures and adapt to environments or tasks. They can be homogeneous, with identical modules, or heterogeneous, with different module types. Examples of modular robots include the CKBot chain architecture system, the Atron lattice system, and the M-Tran hybrid system. Modular robots can reconfigure deterministically by direct manipulation or stochastically by random motion and have applications in areas like space exploration, disaster response, and customizable devices.

1. Configurable Robots
S Austin Moses

2. Modular Self-Reconfiguring (MSR) Robots
• Robots composed of a large number of repeated modules that
can rearrange their connectness to form a large variety of
structures.
• They are able to deliberately change their own shapes by
reorganizing the connectivity of their modules to adapt to
environment or perform new tasks.
• Best known example of an MSR robot would be the
fictional T1000 liquid-metal robot from the film, Terminator
2: Judgment Day.
• Self-reconfigurable mechanism utilizes two types segment
articulation
– Lattice reconfiguration
– Chain reconfiguration.

3. Architecture
• Chain Architectures:
– Modules are connected together in
a string and tree topology.
– Motion controls of the modules are
executed sequentially.
– Easier to design and implement.
• Lattice Architectures:
– Modules are connected in some
regular, space filling 3 D pattern.
– Control and motion are executed in
parallel.
– More flexible and efficient to form
complex structures.
– More suitable for dynamic
environments.

4. Homogeneous and Heterogeneous
Reconfigurable Robots
Homogeneous
• All modules are all the same
• Module position determines role
• Less-costly hardware/software design process
• Simple to reconfigure
Heterogeneous
• Can have different modules
• Function of module determines role
• Many different hardware/designs – costly
• Complex reconfiguration

5. Chain: CKBot System
• The CKBot system is a reconfigurable
robotic system developed by Yim at
the University of Pennsylvania.
• These modules utilize a servo to
rotate one portion of the module with
respect to the other.
• Global inter-module communication
through CANbus as well as local
neighbor-to-neighbor communication
is incorporated on the modules. This
system has also been used in some
experiments in self-repair.

6. Lattice: Atron System
• The Atron system reconfigures in a lattice
system, but can form chains as well. This
image shows a four “legged” or “wheeled”
configuration depending on how the
modules are actuated.
• Modules can distribute power via their
bonding mechanisms and use a power
management system for voltage regulation
and battery charge maintenance.
• A module consists of two hemispheres
where one can rotate continuously relative
to the other.
• Reconfiguration is performed by having one
module grab another and then rotate some
multiple of 90 degrees to another position in
the lattice structure.

7. Hybrid: M Tran System
• The M-TRAN system developed by Murata at AIST/Tokyo Institute of
Technology combines the positive capabilities of chain and lattice based
systems to implement a highly maneuverable and reconfigurable system,.
• A module consists of one passive and one active cube that can pivot
about the link that connects them and can form chains for performing
tasks.
• However during reconfiguration, each of a module’s two cubes can
occupy a discrete set of positions in space when attempting to align with
another module and bond for reconfiguration as in a lattice system.

8. How Do They Re-configure ?
• Deterministic reconfiguration:
– Relies on units moving or being directly manipulated into their target
location during reconfiguration.
– The exact location of each module/unit is always known.
• Stochastic reconfiguration
– Relies on units moving around using statistical processes.
– The exact location of each unit only known when it is connected to the
main structure, but it may take unknown paths to move between
locations
Self-reconfigure
• The units can do this in three ways
– They can move among each other.
– They can change their size.
– They can change a particular property, like color.

9. Complexity in Robot Configuration
• With numerous configuration for a set of modules, the problem of
recognizing & choosing useful config. is a central area of research.
• Factors- Processor organization(centralized/decentralized), inter-module
communication schemes(global bus, local neighbour-neighbour),
module labelling(unique module id vs unlabeled) and
structural symmetry all add to modular systems complexity.
• Centralized-With help of identifying labels central controller
designates explicit commands over global bus. There is no indication
of relative location of modules within the configuration.
• Global & Neighbour-neighbour – adjacency is not enough to represent
full kinematics relationship between 2 modules.

10. Approaches – Robot Configurations
• Variants of adjacency matrices that take into account how structures
are put together & adds essential structural information, such as inter
module-port connections.
• When doing self discovery its often useful to see if a configuration is
same as another configuration by matching a configuration with the
one in a library of configuration.
• In case of neighbour to neighbour communication(ATRON),
distributed algorithms that employ processors of modules interacting
together in parallel, divides the computation required. These MSR
systems use token type messages to pass configuration info from
module-module.
• Enumeration algorithm developed by chen more precisely counts the
number of non-isomorphic configurations.

11. APPLICATIONS
• Studies of the flow of excitation in heart tissue, the dispersal of
medicinal drugs, and pattern recognition
• Space - If you send a robot to Mars, for example, and it breaks, there
is little you can do. But if instead of sending a fixed robot you send a
robot with a supply of modules, then that robot may be able to self-repair
and they need to sustain operation for long periods of time
without human assistance.
• Search & Rescue - Disaster areas such as those around collapsed
buildings or other structures present another type of highly
unstructured unpredictable environment where the use of an MSR
robot could be beneficial. For example, the MSR system could take
the form of a snake which can more easily squeeze through small void
spaces to find victims.
• Bucket of Stuff - The system would be a consumer product comprised
of a container of reconfigurable modules that would reconfigure to
accomplish arbitrary household tasks. This application can be seen as
the most general practical goal of MSR robotics: a system that can
adapt to any task in real time

12. References
• MARK YIM, PAUL WHITE, MICHAEL PARK, JIMMY SASTRA
(2009) Modular Self-Reconfigurable Robots. In: Encyclopedia of
Complexity and Systems Science, University of Pennsylvania,
Philadelphia, USA, pp 19–32