7B-3-4 
for these factors including the increasing complexity of 
production plants, the growing width and depth of au 
tomation, increasing demands with regards to the qual 
ity of the products and security and environmental pro 
tection. Additionally time dimension’s another concern 
as there is an effect on the delivery of products due 
to the decrease on warehouse capacity (Just-In-Time- 
production). 
If a production plant stands still the delivery pro 
cess could stagnate. Despite the separate measures of 
the maintenance process a higher process is responsible 
for organisation and planning. In this process the appro 
priate maintenance strategy, analysis of measures and 
the necessary working plan will be acquired. According 
to Zutt & Hubig [1998] the maintenance process can be 
seen as a circuit (Fig. 2). 
The planning process plays a very important role 
in this circuit. Results, damages and disturbances must 
be checked and analysed to determine the planning of 
the maintenance process. The shortening or even avoid 
ance of machine-standstill’s due to damages, is one of 
the main goals of maintenance. A preventive reduc 
tion of standstill’s, e. g., through intensive inspections 
of machine activities, only make sense if it is economi 
cal sound. Economic viability of production can be im 
proved through efficient maintenance or through effi 
cient planning of the inspection process. 
Considering these points we can assume that the 
maintenance process within a production process is of 
great importance. The increase of efficiency and eco 
nomic viability of maintenance processes with the help 
of modern information-, communication-, and computer 
technologies is necessary and an achievable task. 
3.2 Objectives 
Computer, worn like clothes (wearable computer) com 
bined with wireless communication techniques for the 
interaction between the mobile computer and server 
station, help to optimise the maintenance process of 
big technical plants. Extensive data analysis and dis 
covery of new knowledge, provide valuable information 
for the upcoming planning activities. Research within 
the area of knowledge discovery (KDD) has become vi 
tal over the last few years. The objective is not to build 
a theoretical environment, but to implement new in 
formation technology with the help of new hard- and 
software methods. The outcome will be a highly suit 
able infrastructure which can be used by production 
plants directly. See figure 3 for an overall example. 
In order to achieve this goal it is important to have 
an industry partner who’s maintenance processes takes 
a significant amount of. their annual budget. Our part 
ner is a steel factory and concepts will be developed 
for their 127 industry cranes. The strategy for these 
cranes is a no-mistake-strategy, which means that loss 
of function due to technical faults should be avoided. 
The technical stage of these cranes has to be deter 
mined thoroughly so that critical changes can be de 
tected early. The inspection process is manual at the 
moment and is connected to administrative work. One 
of the consequences of this is, that there is little time 
for the analysis of the inspection results. This is an im 
portant process and therefore we focus on this section 
of the maintenance process. 
Wearable computer, computer which can be worn 
at the body hands free, have the potential to use infor 
mation technology in new situations. This technology 
is beginning to penetrate the market. The inspection 
of cranes, while the production process is running (not 
only because of security reasons) means that the staff 
have hands free to hold themselves. Due to this, no 
electronic data acquisition solution has been created to 
date. Wearable computers can support the data acqui 
sition process directly and efficiently 
Sometimes, this technique is only viable within ap 
plications that can communicate with a basis station 
interactively, e. g. if the user of the mobile station can 
access resources from the server. For example: 
- Access of additional information, necessary for the 
actual situation. 
- Server systems with massive databases. 
- Computing power which ’’overlaps” the mobile sys 
tem. 
In situations where unrestricted mobility is the pre 
mise only the combination of wearable computer and 
wireless LAN can be considered (local, wireless net 
works in areas such as production halls or work premises 
as well as mobile radio networks for more general pur 
poses, from GSM to satellite systems). Applications 
running on a wearable computer must react flexible, 
on the available bandwidth, without leaving the well 
understood interface level to Internet technologies. In 
addition, mobile conference systems with their control 
components have to consider special conditions of the 
connection of radio networks, e. g. varying bit-rates with 
barriers. 
One objective of this project is to provide a work 
ing communication platform for interactive use within 
the inspection process. In order to achieve this goal we 
have to prove how to communicate in difficult environ 
ments, e. g. in a production hall out of steel. Within the 
communication between mobile computer and server it 
is important, if not essential to reduce the amount of 
data being transferred. If we use GSM with a transfer 
rate of max. 14.4 kbit/s (HSCSD) for example, it is 
necessary to prepare and reduce the data transferred 
such that only the ’’important” content for the actual 
query transferred. Data mining (DM) techniques, such 
as Boronowski’s multiple time series analysis [Boronow- 
ski, 1998] is being used.