SSPI: Scalable search of product life cycle
information
This work is funded by SSF.The goal of the SSPI project is to
develop software systems for efficient and scalable search of product
data and meta-knowledge produced during the entire product life cycle.
It is intended to develop such software systems based on Data Stream
Management Systems (DSMS) and the semantic web model.
The project is expected to produce (1) novel methods for scalable
search in high volume data streams, (2) novel scalable methods to
verify systems based on DSMS technology, (3) novel searchable semantic
web based models to preserve product data, and (4) novel ways to verify
the functioning of products in use. The proposed project is
further expected to produce software prototypes which demonstrate the
feasibility of the developed approaches, software which in the hands of
industrial partners may significantly improve their possibility of
developing and selling functional products with increased service
content.
Searching product data
As the volume of streaming data from products in
use may be large, there is need for highly scalable DSMS technology to
provide fast search in these streams. Furthermore, finding the relevant
data streams require suitable meta-data representations and it is
expected that semantic web representations such as RDF and
RDF-Schema as very promising future knowledge representation languages
for such data stream identification.
Product Modelling
and Simulation
The use of simulations in product development is
continuously
increasing, leading to an increased need of validation activities. In
order to validate simulation models, measured real world data is
needed. In the SSPI project it will be investigated how simulation
models can be continuously verified by using a DSMS. The verification
is made by continuously comparing measurement streams from products in
use with the predicted behaviour generated by the simulation models. In
case of deviations, the deviations may either indicate faulty behaviour
of the product or incomplete or erroneous simulation models. In the
latter case the simulation model needs to be further developed.
Semantic Web Model
Approach
Product models document product information
along with the knowledge used by the people developing the product (for
example design rationale). Knowledge about why something is designed a
certain way (knowledge about knowledge) is called meta-knowledge.
Efficient and scalable preservation, reuse, and search in
meta-knowledge is of utmost importance in product developments.
Long term preservation of product meta-knowledge and data requires
representation languages that can describe information from regular
databases, from product models, and from (archived) streams produced by
products in use.
In the SSPI project it will be investigated how
to preserve and search all different kinds of product information based
on semantic web based meta-knowledge.
Project organization
Responsible for this project is Tore Risch. It is a collaboration with Lennart Karlsson at Luleå University of Technology.
© 2009 Uppsala Universitet, Department of Information Technology, Box 337, 751 05 Uppsala, Sweden | This page is maintained by Tore Risch