Definition of the warranty cases limitation and modifying the excel

This thesis only handles the Factory 3’s warranty reports between xx and xx. After defin-ing that the actual product models, which this thesis is based on, would include only the current volume production models without the obsolete or project models, the final war-ranty report amount was delimited to xxx warwar-ranty reports. This claims amount is too massive to handle without a helping tool. Happily, the CFC report tool aided considera-bly, in handling the warranty collection in excel tables, which included most of the main information from the reports.

The time limit’s closing date was xx, due to the schedule of this thesis. However, the total case amounts of xx are evaluated by approximate values, making the comparison of the year xx possible with the previous years.

The research of this thesis is based on an excel data sheet driven from the CFC data bank.

This excel data sheet was exported from the CFC data bank with the CFC report tool. The excel data sheet includes basic information on all of the CFC warranty reports, which is mainly listed below:

 Initiated by – the person who has created the CFC document

 Product Group

 Product Name

 Machine serial number

 CFC number

 CFC Subject – excel table’s main information about the issue Whitewashed text.

Obviously the excel table didn’t include the warranty claim’s text explanation about the issue, any photo from the site, or any additional attachments for wider justifications for the case. All of this information can be mechanically found from the warranty claim, but it takes some time, and is not very handy compared to reading the information from the excel columns. Because the warranty reports’ quality varies a lot, it’s possible that the excel table has nearly all of the information on some smaller cases.

The original raw excel data had challenges with data quality. The Module Name –column, which was essential for obtaining the information to divide the warranty cases into dif-ferent groups, posed the biggest challenge. It had enormous amount of difdif-ferent options and spelling types, for example sensor, Sensors, Sensor, electric, Electrics, Electrifica-tion, Electrifications, Steel structure, Steel structures, Other, blank, ?, -, 101010 and so on, with around two hundred options. However, misspelling wasn’t the biggest challenge, but the totally incorrectly chosen Module Names. In some cases the Module Name was typed electrics, but the actual problem was with hydraulics. This caused mistrust for the data, and forced the processing of a huge amount of warranty claims, which was a labo-rious project, but had to be done to get the wanted results from this research.

Shifting through the raw warranty claims data began by fixing the module categorization to get information on which modules fail, from which machine models, and how much costs it has inflicted. The module categorization is based on the products’ actual structure;

therefore both crushing and screening lokotracks’ final assembly is made from separate modules, such as frame module, feeder module, crusher module and conveyor modules.

This original module categorization is based on the product design and PDM system, but it has been tuned to fit the best possible way for this thesis’ topic, and uncategorized raw data. In addition, electrics and hydraulics are separated from the original product’s struc-ture, because they can be better managed on their own. Module categorization is more useful, when some of the similar kinds of parts are merged into one – for example, whole hydraulics and electrics categories, regardless of in which module they are attached.

While sorting out the Module Name –column, the Sub Module -column was added to improve the data for deeper analysis results. First, the original xx different module names were reduced to 12 different modules, which are shown above, and these modules were again divided into separate sub modules, so now the 12 modules have a total amount of xx different submodules, around x to y for each module. Modules and sub-modules to-gether with the other data columns will give different alternatives to categorize a large amount of warranty reports, and to compare the quality levels between the products or product groups by modules and sub-modules, or to separate total costs from possible quality epidemics.

Categorizing the warranty claims by submodules wasn’t easy or clear work to do, and although it has been done by professional skills about the machine models, some com-promises had to be made to keep the amount of sub-modules reasonable. Different prod-uct models and prodprod-uct categories technically vary from each other, which had to be ac-commodated, when adding a sub-module column to the excel sheet. Now sub-modules are mainly comparable with products or product categories, and thus they can be com-pared and analyzed.

Final categorization by modules is proposed in the following list:

Whitewashed text.

This categorization is based on the product structure. Example product LT is shown in figure below.

Figure 11 – Example LT (brochure)

An example on sub-module categorization is presented in the list and figure below. xxxx module is a good example on sub-module categorization:

Example module: Conveyor

 Bearing

 Belt

Figure 12 – Example conveyor structure

As shown in figure above, dividing the warranty claims by submodules is based on the product structure. In this example, the conveyor module is divided into 8 separate sub-modules. Besides these submodules, categories multiple and other needed to be added.

These two categories summarize all of the other warranty claims set under conveyor mod-ule – “multiple” are those warranty claims which have been created from multiple reasons that all have to do with the conveyor. Under the submodule “other” are sorted the war-ranty claims that were not possible to categorize into the 8 separate submodules presented above. However, these two non-specific submodules were obligatory to create, and these won’t impact the thesis’ conclusions, because they usually are only a fractional part of the whole claim amount, or represent shares of other categories, and consequently won’t skew the conclusions. Similar submodule-categorizing with similar kinds of compromises has been done under all of the xx upper level modules. These were obligatory due to the raw data quality.

In addition to these categorizations, all of the machine models in this thesis are also listed under five product groups: B-series barmac crushers, C-series jaw crushers, GP-series cone crushers, LT-lokotracks and ST-lokotracks. This categorization is based on the prod-uct categories which are widely used in marketing and prodprod-uction over the whole com-pany and by customers, and it also gives some depth for the analysis further on this thesis.

When a customer machine has a breakdown issue which affects the customer’s production processes, the main goal of service is to repair the machine as soon as possible to let the process flow again. These machines are always a part of some bigger aggregate, and the breakdown causes disturbance to the whole process. After repairing the machine, (by fix-ing the broken part, or by replacfix-ing the part with a new one), the root cause analysis is usually quite difficult to execute, and the proper analysis work should be at least started before or while reparation. Unfortunately, the root cause analysis is often done just with the help of pictures, which causes some uncertainty to the effectiveness of the corrective actions.

Other aspect making categorization a bit complex was the total amount of claims. It took a significant time to take a look at the claims from the database and go through the excel table. Categorizing the warranty claims by modules was necessary for driving graphs about the issues on the upper level, but still more detailed, than the bare machine model allows. Furthermore, determining different subcategories for each separate upper module gave the opportunity to receive more specific information on costs’ formation.

In addition, warranty reports are categorized by responsible authors. However, this cate-gorization is not as distinct as the one by machine models, broken modules and submod-ules or by operating time.

Whitewashed text.

In this thesis the choice was made, that a good way to settle the warranty case amounts, and to improve product quality is to divide the warranty cases into different categories.

Afterwards the analysis of the results the problems should be solved as interdisciplinary teams consisting of team members from engineering, manufacturing, purchasing and quality management, connecting with the suppliers, service men and customers as needed.

In addition, for example, the manufacturing and sales volumes have been added to give weight to the warranty data and analysis.