Focusing Data Transformation in Product Lifecycle ManagementMay 14, 2018
Traditionally PLM consulting has focused only on People, Process and Technology. Data was not even considered even relevant in the consulting paradigm. This view is now changing very fast and data is now increasingly playing a substantial role in providing analytical decision-making capability to cross functional teams. We propose that PLM consulting should comprise of people, process, technology and data.
Data is increasing every second at an exponential rate. Every action we take generates data. PLM is no exception. Each and every stage in a Product Lifecycle Management (PLM) process generates enormous data.
“As per current trends it is expected that the world business data will be doubled every 1.2 years”
This paper will present the business value data brings during a PLM transformation process in any organization. Data can be used to provide useful business insights. They can enable cross functional teams take PLM process decisions with confidence. Cost reduction, supplier rationalization, attribute rationalization and process optimization as just some of the huge benefits proper data utilization brings to us.
Understanding Data in PLM
What data are we talking about?
The Oxford Dictionary defines data as –
“Facts and statistics collected together for reference or analysis”
Data is generated throughout the product lifecycle. It may be in form of information collected at these lifecycle steps or external input that is provided at a PLM phase. Most of the information is collected digitally these days, while some information is still collected as hard copy documents. Most of the information collected today in PLM systems are standalone. They are almost never analyzed for process efficiencies or for other product benefits. They are just meant to adhere to a predefined PLM process.
Different types of Data in a typical organization
There are essentially three different types of data in an organization –
– This refers mainly to the data that is stored in a database in an organized manner. Other sources of structured data can be excel spread sheets, flat files, csv files etc. It is easiest to read and manage structured data. End vision of each organization should be to reach 100% structured data.
Semi Structured Data
– This refers to data that is somewhat structured, but still has some unstructured elements in it. It is not possible to directly read a semi-structured data. However standard techniques exist to read, manipulate and write structured data. Some examples include XML files, HTML files, log files etc.
– It refers to data that is disorganized and not ready for direct processing. The data is not stored in any proper order or layout. It is the toughest to extract, convert and transform unstructured data to structured data. Currently there is no standard way to perform this activity.
Challenges without Data Transformation
Raw data can be consumed for analysis only after the entire dataset is converted into a standardized digital form. Data transformation is challenging in many aspects. One cannot proceed with PLM Data Transformation without addressing them.
“Through 2019, more than 50% of data migration projects will exceed budget and/or result in some form of business disruption due to flawed execution.”
The challenges during a typical Data Transformation process can be summarized by the below topics—
The volume of data is growing exponentially every moment. It more than important to build a scalable solution that can handle the increase in data. Data organization becomes a huge challenge with high volume of the data. If the data is not organized properly then data reusability and searchability are impacted. This leads to either creation of duplicate data or people start working using wrong data sets.
Distributed Data Sources
Like any other field, IT had also taken PLM by storm. This led to creation of multiple disintegrated data sources for large enterprises. These sources can vary from local hard drives to data sharing application. This makes accessing and finding the data very difficult. Also, this can lead to data synchronization problem
Not all data in an organization is available digitally. Plus digital data also has several forms for example – text, documents, pdf, database etc. It is challenging to integrate and channel them for bringing value to a PLM system.
We analyzed a leading life sciences organization for data format variation and found out that the specification data is currently stored in currently 24 different formats. This implies that the non-standardized templates were used in creation of these documents, leading to creation of different document formats. This poses a huge business risk in terms of noncompliance. Not all document formats are reusable and hence causing a lot rework.
Duplicate data can lead to serious consequences in an organization. To say the least, it wastes useful resources for an organization. A small change in the data must be reflected at all places. Even if one source is not updated then risk of people working on wrong data grows exponential. It can have ramifications like incorrect costing, compliance risks and incorrect inventory management. Thus, proper identification and elimination of duplicate data is critical before proceeding with data transformation.
Regional Influence on Data
Data too has regional influence. People in many countries store vital business data in their own native languages, making them inaccessible to the global organization. This leads to a lot of rework and waste of useful resource.
In a case study we found that 41.5% of unstructured data in a Global Consumer Organization was in native language. Data in 18 different langauges had no or very little global translation available. This made the data unusable globally, leading to a lot of rework and loss of precious information.
Identification of Stale Data
Improper data archiving strategy often leads to bulky systems where important data is very had to find. A lot of irrelevant data that is not longer required is just sitting in the system. A proper data achiving strategy is required to deal with obsolete data.
Benefits of Data Transformation in PLM
Benefits of Data Transformation Program in a PLM transformation phase can have far reaching effects. It can not only aid the PLM transformation program by providing critical analytical data needed for PLM process optimization, it can also touch other improvement areas as well. Some of the benefits of Data Transformation in PLM are discussed below.
Enterprise search is perhaps the most used feature of a PLM software. The efficiency of the search depends on the nature of data present in the system. During data transformation we extract, transform and load the unstructured data in a structured manner. This unstructured data that was once not searchable is not present as metadata in the system. This makes is easier for engineers and designers to search objects in the PLM system. This reduces the chances of duplicate part creation, saves the precious resources time and hence aids faster profuct development.
Aids in Supplier Assessment and Rationalization
Large organizations have huge unstructured datasets. These organizations have been dealing with suppliers and vendors thoughout geographies. They have large supplier documents that are lying idle in the file storage systems. Using data transformation, we can extract supplier metadata from these unstructured documents. This metadata can be used to compare suppliers for raw materials, cost, location benefits etc. Data Transformation can thus position us in a better bargaining position and help us drive more value in our interaction with vendors and suppliers.
Identify hidden data connections and support PLM Process Optimization
It is very difficult to identify product linkages in unstructured data. One has to manually scan all the documents to check if there is any product linkages. After data transformation we automatically search the product metadata to see if there are any linkages with other products. This will enable us to determine the impact of making a change in a single product. In another case we can predict the ideal bill of materials or bill of specifications by analyzing linkages in transformed data. This will help cross functional process teams to make PLM designs with confidence.
Improve Process Complexities
It can be very challenging in large organizations to optimize a PLM process. Data Transformation of email data into a structured metadata can help identify process complexities automatically. We can easily identify the critical thread, overloaded resources, and non-productive transactions using Data Transformation. This will lead to huge cost and time benefits.
Better Data Governance
Centrally organized structured single source of truth can enable organizations to function more effectively. After data transformation this single source of truth can be made more effective. Since bulk of the enterprise data resides as unstructured content, with data transformation we can make a lot of product data available structurally. This can help in sending accurate information to suppliers, to manufacturing units, to internal team, to government agencies and to external clients.