Developing and Maintaining an Ontology for Rehabilitation Robotics
Zeynep Dogmus, Gizem Gezici, Volkan Patoglu and Esra Erdem
Faculty of Engineering and Natural Sciences, Sabancı University, Istanbul, Turkey
Keywords:
Rehabilitation Robotics, Ontology Development.
Abstract:
Representing the available information about rehabilitation robots in a structured form, like ontologies, fa-
cilitates access to various kinds of information about the existing robots, and thus it is important both from
the point of view of rehabilitation robotics and from the point of view of physical medicine. Rehabilitation
robotics researchers can learn various properties of the existing robots and access to the related publications
to further improve the state-of-the-art. Physical medicine experts can find information about rehabilitation
robots and related publications (possibly including results of clinical studies) to better identify the right robot
for a particular therapy or patient population. Therefore, considering also the advantages of ontologies and
ontological reasoning, such as interoperability of various heterogenous knowledge resources (e.g., patient
databases or disease ontologies), such an ontology provides the underlying mechanisms for translational phys-
ical medicine, from bench-to-bed and back, and personalized rehabilitation robotics. With these motivations,
we have designed and developed the first formal rehabilitation robotics ontology, called REHABROBO-ONTO,
in OWL, collaborating with experts in robotics and in physical medicine. We have also built a software (called
REHABROBO-QUERY) with an easy-to-use intelligent user-interface that allows robot designers to add/modify
information about their rehabilitation robots to/from REHABROBO-ONTO.
1 INTRODUCTION
Ontologies (like databases) are formal frameworks for
representing knowledge in a structured form, to aid
access to relevant parts of the knowledge and auto-
mate reasoning over it. An ontology can be viewed as
a graph where nodes denote concepts (e.g., rehabilita-
tion robots, joint movements) and the edges between
the nodes denote relations between the corresponding
concepts. For instance, an edge from a node that de-
notes “Upper Extremity Rehabilitation Robots” to a
node that denotes “Rehabilitation Robots” may char-
acterize the “is-a” hierarchy relation; whereas an edge
from a node that denotes “Rehabilitation Robots” to
a node that denotes “Joint Movements” may char-
acterize “targets” relation. Due to their flexible
graph-like structure, ontologies (unlike databases) al-
low representation of incomplete knowledge, can eas-
ily be extended by new information (e.g., with new
sorts/features of rehabilitation robots). Due to their
formal representations, ontologies developed by dif-
ferent parties at different locations can be integrated,
and reasoning (e.g., query answering) can be auto-
mated over concepts and their relations represented
in these ontologies. Therefore, it is not surprising that
more and more knowledge-intensive systems (includ-
ing Semantic Web (Berners-Lee et al., 2001) that is
planned to provide automated services to Web by giv-
ing meaning to concepts) rely on ontologies to enable
content-based access, interoperability, and communi-
cation across the Web.
As the number of rehabilitation robots increase,
the information about them also increases, but most of
the time in unstructured forms (e.g., as text in publi-
cations), which make it harder to access the requested
knowledge (e.g., the flexion/extension range of mo-
tion (RoM) of ASSISTON-WRIST (Erdogan et al.,
2011)) and thus reason about it (e.g., finding the reha-
bilitation robots that target shoulder movements and
also have at least 60
RoM for the flexion/extension
movements of the wrist). To facilitate access to re-
quested knowledge about rehabilitation robots, we
have designed and developed the first formal rehabili-
tation robotics ontology, called REHABROBO-ONTO.
This ontology has been designed in a way that en-
ables integration with other medical ontologies, such
as ontologies that capture rehabilitation protocols,
patient data and disorder details. Considering the
standards of World Wide Web Consortium (W3C),
REHABROBO-ONTO is represented in OWL (Web
Ontology Language) (Horrocks et al., 2003; Antoniou
and van Harmelen, 2004).
389
Dogmus Z., Gezici G., Patoglu V. and Erdem E..
Developing and Maintaining an Ontology for Rehabilitation Robotics.
DOI: 10.5220/0004145303890395
In Proceedings of the International Conference on Knowledge Engineering and Ontology Development (KEOD-2012), pages 389-395
ISBN: 978-989-8565-30-3
Copyright
c
2012 SCITEPRESS (Science and Technology Publications, Lda.)
Our goal is to make this ontology open-source so
that every rehabilitation robotics researcher can eas-
ily add information about his/her robot to it, and ev-
ery rehabilitation robotics researcher and every phys-
ical medicine expert can access information about
all available rehabilitation robots. To facilitate such
modifications and uses of REHABROBO-ONTO, we
have developed a software (called REHABROBO-
QUERY) with an intelligent user-interface. In this
way, experts do not need to know logic-based repre-
sentation languages of ontologies, like OWL, or Se-
mantic Web technologies, for information entry, re-
trieval and modification.
The ontology system consisting of REHABROBO-
ONTO and REHABROBO-QUERY is of great value
to robot designers as well as physical therapists and
medical doctors. On the one hand, robot designers
can benefit from the system, for instance, to identify
robotic devices targeting similar therapeutic exercises
or to determine systems using a particular kind of
actuation-transmission pair to achieve a range of mo-
tion that exceeds some threshold. Availability of such
information may help inspire new designs or may lead
to a better decision making process. The ontology can
also be utilized to group similar robots by quantifiable
characteristics and to establish benchmarks for sys-
tem comparisons. Overall, an ontology designed to
specifically meet the expectations of the overall reha-
bilitation robotics effort has the potential to become
an indispensable tool that helps in the development,
testing, and certification of rehabilitation robots. On
the other hand, physical therapists and medical doc-
tors can utilize the ontology to compare rehabilitation
robots and to identify the ones that serve best to cover
their needs, or to evaluate the effects of various de-
vices for targeted joint exercises on patients with spe-
cific disorders.
It is important to emphasize that the ontology
REHABROBO-ONTO and the tool REHABROBO-
QUERY introduced in this paper have been developed
to initiate efforts in utilizing ontological technologies
for the field of rehabilitation robotics. Therefore, by
making REHABROBO-ONTO available open-source
via REHABROBO-QUERY, it is our intention to con-
tinually update and enhance capabilities of these tools
according to the feedback provided by the commu-
nity.
2 RELATED WORK
Although there are some ontologies maintaining in-
formation about objects or environments (Chella
et al., 2002; Yanco and Drury, 2004; Paolucci and
Sycara, 2004; Wang et al., 2005; Suh et al., 2007;
Johnston et al., 2008), developed for the use of robots,
there are only several works in the literature that have
proposed ontologies about robots.
In particular, Amigoni and Neri (Amigoni and
Neri, 2005) introduce two ontologies (in OWL):
one to store general concepts and properties/relations
about the movement capabilities of mobile robots
(e.g., wheels and their properties) and the other to de-
scribe the high level tasks that these robots can per-
form (e.g., move, rotate). The idea is then to allocate
tasks and/or assign roles to mobile robots by means
of querying these two ontologies using a description
logics reasoner.
Schlenoff and Messina (Schlenoff and Messina,
2005) introduce an ontology (in OWL) for urban
search and rescue robots. The ontology captures
structural characteristics (such as size), functional ca-
pabilities (such as locomotion capabilities) and oper-
ational considerations (such as display type) of the
robots with a goal of assisting in the development and
testing of search and rescue robot systems.
Juarez et al. (Juarez et al., 2011) introduce a
database (called ROBODB) for storing physical char-
acteristics of robots; but also note that they plan to
transform the knowledge stored in ROBODB into an
OWL ontology to benefit from this “common” lan-
guage of ontologies and related reasoners.
However, none of these existing robot ontologies
have been designed to target rehabilitation robots and,
without further customization, they fail to capture
many important aspects of rehabilitation robots, in-
cluding the interoperability with the existing ontolo-
gies in physical medicine.
3 DESIGNING AN ONTOLOGY
FOR REHABILITATION
ROBOTS
We have designed the rehabilitation robots ontol-
ogy REHABROBO-ONTO considering suggestions of
the rehabilitation robotics researchers and physical
medicine experts whom we collaborate with. As sug-
gested in (Uschold and King, 1995) about design-
ing an ontology, we have first identified the purpose,
and then identified and defined the basic concepts and
their thematic classes, and their relationships for the
chosen subject domain.
Our goal of developing an ontology for rehabilita-
tion robotics is mainly to maintain a knowledge repos-
itory containing information about all rehabilitation
robots and relevant references, to facilitate access to
KEOD2012-InternationalConferenceonKnowledgeEngineeringandOntologyDevelopment
390
Figure 1: REHABROBO-ONTO with main classes.
requested information in this repository for both robot
designers as well as physical medicine experts. In this
way, not only it will be easier for robot designers to
improve the state-of-the-art in rehabilitation robotics
but also it will aid translation from bench-to-bed and
back, and personalized physical medicine by allow-
ing the physical medicine experts to choose the right
rehabilitation robots for specific patients/therapies.
Along these goals, next comes the question of
what kind of information should be maintained in
such an ontology about rehabilitation robots.
We have designed our ontology (Figure 1) consid-
ering four main concepts (or thematic classes):
RehabRobots (representing rehabilitation robots
and their properties),
JointMovements (representing targeted joint
movements and their properties),
Owners (representing robot designers who
add/modify information in the ontology about
their own robots),
References (representing publications related to
rehabilitation robots).
These concepts are related to each other by the fol-
lowing relations:
a rehabilitation robot targets joint movements,
a rehabilitation robot is ownedBy a robot designer,
a rehabilitation robot hasReferences to some
publications.
As seen in Figure 1, each class has its own prop-
erties. For instance, RehabRobots has the “func-
tionality” property to be able to describe that a re-
habilitation robot can be used at home or in clinic.
JointMovements has a property to describe the
“RoM type” of the robot: a rehabilitation robot can
be used in active mode, where some effort from the
patient is required while the robot guides the patient,
or in passive mode, where no effort is required from
the patient while the robot guides the patient’s move-
ments. We also keep information about the owners
(e.g., names, institutions) and the related references.
Considering various sorts of rehabilitation robots
and various sorts of joint movements, RehabRobots
and JointMovements classes have subclasses; some
of these subclasses are illustrated in Figures 2 and 3.
Maintaining such a hierarchy aids not only com-
pact representation of knowledge about rehabilitation
robots (by avoiding repetitions) but also efficient rea-
soning about it. Currently there are 93 classes repre-
sented in REHABROBO-ONTO.
DevelopingandMaintaininganOntologyforRehabilitationRobotics
391
Figure 3: Hierarchy of lower extremity joint movements targeted by rehabilitation robots.
Figure 2: Hierarchy of lower extremity rehabilitation
robots.
4 DEVELOPING AND
MAINTAINING
REHABROBO-ONTO
Ontologies represented formally in a language, such
as OWL, are based on variations of Description Log-
ics (DL) (Baader et al., 2008). DL provides the logi-
cal formalism not only for such formal ontologies but
also the Semantic Web.
DL terminology consists of concepts, roles, and
objects. Objects denote entities of our world with
characteristics and attributes; concepts are interpreted
as sets of objects; and roles are interpreted as bi-
nary relations on objects or concepts. According
to a formal ontology terminology (e.g., in OWL),
concepts are called classes, attributes of classes are
called data properties, roles are called object prop-
erties, and objects are called individuals. For in-
stance, a concept/class named RehabRobots may
represent all rehabilitation robots, whereas an ob-
ject/individual named ASSISTON-SE represents the
particular shoulder robot introduced in (Ergin and
Patoglu, 2012; Yalcin and Patoglu, 2012). A
class (e.g., RehabRobots) may have subclasses (e.g.,
KEOD2012-InternationalConferenceonKnowledgeEngineeringandOntologyDevelopment
392
ShoulderRobots); then subclasses inherit properties
of classes. Such a possibility of representing hierar-
chical classes allows compact representation as well
as efficient reasoning over it.
Due to the common logic-based formalism that
underlies formal ontologies and Semantic Web,
different ontologies developed by different parties
around the world can be integrated for deep auto-
mated reasoning. In that sense, considering also the
possibilities of integrating information about rehabili-
tation robotics with other information (e.g., patient in-
formation, disease information, genetic information),
maintaining information about rehabilitation robotics
as formal ontologies is well-decided.
Once we identify classes, their properties, sub-
classes, relations between classes, and define them
precisely as in the previous section, we can repre-
sent the ontology formally using a logic-based ontol-
ogy language. Considering the standards of W3C,
we have decided to represent the ontology in the
logic-based ontology language OWL; then we need
to use description logic reasoners to find answers
to experts’ queries. We have represented the OWL
ontologies that describe general concepts and their
properties/relations using the OWL ontology edi-
tor PROT
´
EG
´
E (Gennari et al., 2003).
Once we represent general concepts and their
properties/relations about rehabilitation robots in
OWL, we are not done yet. As discussed above as
a part of our goals, we would like the rehabilita-
tion robotics ontology to be shared by researchers
so that robot designers can add/modify information
about their robots, and both rehabilitation robotics ex-
perts and physical medicine experts can ask queries
over it. Therefore, we would like to allow researchers
to add information about specific rehabilitation robots
by “assertions” (like, “the rehabilitation robot whose
name is ASSISTON-WRIST is a wrist robot and it has
clinic use”) as well.
Such assertions about specific individuals can
be added to REHABROBO-ONTO using PROT
´
EG
´
E.
However, since PROT
´
EG
´
E downloads the whole on-
tology to be able to add new information, ensur-
ing that the users add information to REHABROBO-
ONTO without letting them modify other parts of the
ontology may be problematic. Also, assuming that
the existing robotics experts and physical medicine
experts know about DL and logic-based ontology lan-
guages, that they have experience in using DL rea-
soners or Semantic Web technologies, and that they
keep track of the most recent versions of these soft-
ware, may not be reasonable along our goals for an
effective use of the rehabilitation ontology. To facil-
itate the effective use of the rehabilitation ontology
Figure 4: An overview of the system architecture of
REHABROBO-QUERY.
by different users, we have designed a tool (called
REHABROBO-QUERY) with an easy-to-use intelli-
gent user-interface.
REHABROBO-QUERY (Figure 4) is a soft-
ware that allows rehabilitation robotics researchers
to add/modify information to REHABROBO-ONTO
about their robots by following consecutive tabs of
the intelligent user-interface and allow experts to ask
queries about the existing robots, without having to
know about DL or logic-based ontology languages,
and without possessing any experience of using the
existing description logic based reasoners or Seman-
tic Web technologies. Some snapshots of the user
interface of REHABROBO-QUERY for adding new
information to REHABROBO-ONTO ontology about
the rehabilitation robot ASSISTON-WRIST (Erdogan
et al., 2011) are shown in Figures 5 and 6.
As the registered user describes the robot by nav-
igating the tabs, the information is held in the system
in a structured way, as temporary objects. Therefore,
the user has the chance to return to any tab to change
the information. After entering all properties of the
robot, in the End tab, all the information entered by
the user is displayed as a summary for the last time.
After the user checks the information and confirms
its addition to REHABROBO-ONTO, the information
about the rehabilitation robot is transformed into as-
sertions in OWL to be added to REHABROBO-ONTO.
Assertions about each rehabilitation robot is stored
in a separate file, to make it easy to modify/delete
REHABROBO-ONTO as well as for efficient query an-
swering planned as part of future work.
A user can modify or delete his/her own robots
only. First the relevant robots are found by querying
REHABROBO-ONTO using the DL reasoner HER-
DevelopingandMaintaininganOntologyforRehabilitationRobotics
393
Figure 5: REHABROBO-QUERY: Adding to REHABROBO-ONTO general information about the rehabilitation robot
ASSISTON-WRIST.
Figure 6: REHABROBO-QUERY: Adding to REHABROBO-ONTO information about the RoM of targeted joint movements of
the rehabilitation robot ASSISTON-WRIST.
MIT (Motik et al., 2007) via OWL API, and listed
in a pull-down menu. For modification, after the user
chooses a robot from the list, the user interface that we
have seen earlier for adding information appears but
now with tabs filled with the robot’s properties. The
user can make changes via this interface and the up-
dated information can be saved as a set of assertions in
OWL, in a new file while keeping the previous version
as “modified”. For deletion, after the user chooses a
robot from the list, the relevant file containing asser-
tions about that robot is marked as “deleted”. Note
that in both cases, we keep the information about the
robot before modification/deletion as well; these files
may be needed if the user accidentally deletes his/her
robot from REHABROBO-ONTO, or modifies it incor-
rectly.
5 CONCLUSIONS
We have designed and developed the first formal re-
habilitation robotics ontology, called REHABROBO-
ONTO, to represent information about rehabilitation
robots. The benefits of having such an ontology can
be summarized as follows:
It provides structured formal representation about
rehabilitation robots and their properties. This
further allows easy access to the requested in-
formation, integration with other knowledge re-
sources (e.g., patient databases, or disease ontolo-
gies), as well as reasoning (e.g., answering com-
plex queries) over all these knowledge resources.
It allows the selection of right rehabilitation
KEOD2012-InternationalConferenceonKnowledgeEngineeringandOntologyDevelopment
394
robots for a particular patient or a physical ther-
apy. In particular, it helps finding available re-
sources (e.g., rehabilitation robots and related
publications that may involve results of clinical
studies) that address a specific need. This further
paves the way for translational physical medicine
(from bench-to-bed and back) and personalized
physical medicine.
It aids exchange of information across rehabilita-
tion robots researchers over the world, and thus to
improve the state-of-the-art.
It allows to identify “gaps” in functionality of re-
habilitation robots, that can further improve re-
search efforts.
We have also built a software, REHABROBO-
QUERY, with an easy-to-use intelligent user-interface
that allows robot designers to add/modify information
about their rehabilitation robots.
Along these directions, our ongoing work involves
extending REHABROBO-QUERY with the function-
ality of answering queries over REHABROBO-
ONTO using a DL reasoner. For instance, using
REHABROBO-QUERY, we will be able to find
the rehabilitation robots that has targeted pop-
ulation “adult” and that targets “shoulder
scapular elevation/depression movements”.
Our ongoing work also includes making
REHABROBO-QUERY available to researchers by a
web-service, and integration of REHABROBO-ONTO
with the existing related ontologies.
ACKNOWLEDGEMENTS
This work has been partially supported by Sabanci
University IRP Grant and TUBITAK Grant 111M186.
REFERENCES
Amigoni, F. and Neri, M. (2005). An application of on-
tology technologies to robotic agents. In Proc. of
IEEE/WIC/ACM IAT.
Antoniou, G. and van Harmelen, F. (2004). Web ontology
language: OWL. In Handbook on Ontologies, pages
67–92.
Baader, F., Horrocks, I., and Sattler, U. (2008). Description
Logics. In van Harmelen, F., Lifschitz, V., and Porter,
B., editors, Handbook of Knowledge Representation.
Elsevier.
Berners-Lee, T., Hendler, J., and Lassila, O. (2001). The
semantic web. Scientific American.
Chella, A., Cossentino, M., Pirrone, R., and Ruisi, A.
(2002). Modeling ontologies for robotic environ-
ments. In Proc. of SEKE.
Erdogan, A., Satici, A., and Patoglu, V. (2011). Passive
velocity field control of a forearm-wrist rehabilitation
robot. In Proc. of IEEE ICORR.
Ergin, M. A. and Patoglu, V. (2012). AssistOn-SE: A
self-aligning shoulder-elbow exoskeleton. In Proc. of
IEEE ICRA.
Gennari, J. H., Musen, M. A., Fergerson, R. W., Grosso,
W. E., Crub
´
ezy, M., Eriksson, H., Noy, N. F., and Tu,
S. W. (2003). The evolution of prot
´
eg
´
e: an environ-
ment for knowledge-based systems development. Int.
J. Hum.-Comput. Stud., 58(1):89–123.
Horrocks, I., Patel-Schneider, P. F., and van Harmelen, F.
(2003). From shiq and rdf to owl: the making of a
web ontology language. J. Web Sem., 1(1):7–26.
Johnston, B., Yang, F., Mendoza, R., Chen, X., and
Williams, M.-A. (2008). Ontology based object cat-
egorization for robots. In Proc. of PAKM.
Juarez, A., Bartneck, C., and Feijs, L. M. G. (2011). Us-
ing semantic technologies to describe robotic embod-
iments. In Proc. of HRI.
Motik, B., Shearer, R., and Horrocks, I. (2007). Optimized
reasoning in description logics using hypertableaux.
In Proc. of CADE.
Paolucci, M. and Sycara, K. (2004). Ontologies in agent ar-
chitectures. In Steffen Staab, R. S., editor, Handbook
on Ontologies in Information Systems. Springer.
Schlenoff, C. and Messina, E. (2005). A robot ontology for
urban search and rescue. In Proc. of CIKM-KRAS.
Suh, I. H., Lim, G. H., Hwang, W., Suh, H., Choi, J.-H.,
and Park, Y.-T. (2007). Ontology-based multi-layered
robot knowledge framework (omrkf) for robot intelli-
gence. In Proc. of IROS.
Uschold, M. and King, M. (1995). Towards a methodology
for building ontologies. In Proc. of Workshop on Ba-
sic Ontological Issues in Knowledge Sharing, held in
conjunction with IJCAI-95.
Wang, E., Kim, Y. S., Kim, H. S., Son, J. H., Lee, S., and
Suh, I. H. (2005). Ontology modeling and storage sys-
tem for robot context understanding. In Proc. of KES
(3).
Yalcin, M. and Patoglu, V. (2012). Kinematics and design
of AssistOn-SE: A self-adjusting shoulder-elbow ex-
oskeleton. In Proc. of IEEE BioRob.
Yanco, H. A. and Drury, J. L. (2004). Classifying human-
robot interaction: an updated taxonomy. In Proc. of
SMC (3).
DevelopingandMaintaininganOntologyforRehabilitationRobotics
395