Let’s look at the numbers involved with tooling a modern Semiconductor Fab

Over 2,000 tools valued at over $2billion, installed by 8,000 contractors from 12 companies over a period of 24 months. Millions of changeable data points flowing through multiple design and construction teams and almost all of it managed with our trusted friend... the spreadsheet.  

A real world case study

Basesite work on Semiconductor projects worth billions of dollars across multiple continents.  A large Semiconductor owner had been managing over 200,000 Points of Connection (PoC)  data across all disciplines in a large facility via a mixture of Excel and hard copy paper.

In the first instance we tabularized the customer data into a database that matched tool PoC data with Facility PoC data. Basesite’s  Tool-Install management software allowed end users to assign Points of Connection within the centralized database for all disciplines. This was aggregated and sortable by tool allowing Project Engineers to view information in real time. Through a couple of clicks Electrical POCs were assigned to available panels with the panel loads, cable lengths and train all automatically aligned and confirmed within design constraints.  

The assignment is instantly propagated throughout the project ecosystem. This compares to the legacy spreadsheet methods where assignments are made in a silo ’ed spreadsheet and then manually verified before being propogated throughout the design ecosystem via email or hard copy paper documents. A process that takes days, means different stakeholders receive the update at different times and is rife with errors as data is duplicated by hand,

We have seen data across multiple projects in multiple countries and continents, it shows that for an average sized tool working via the database solution vs excel spreadsheets, there was an average saving of a staggering 35 hours.  If we extrapolate this out on a small to average 1,000 tool installation facility that’s a saving of 35,000 hours.  

Assuming an average engineering resource cost of $100 per hour the realised cost saving is $3.5 million per facility.  

This base line figure does not include the other benefits that accrue from working with a database :
  1. Elimination of data errors
  2. Time saved from having a live database with multiple stakeholder views meaning changes do not have to be communicated via email/verbally
  3. Custom graphs/reports that can be created to track progress

What's the current status quo?

When managed through a sub optimal process the task of managing all these data points often falls to Project Engineers, after all it is often their data that needs to flow between team across each stage of the construction process.

This leads to high value engineering resources being used to carry out automatable low value tasks – normally managing data points in Excel.  While they are squinting at those spreadsheets, who is looking at designing and detailing.

So, what’s the best method for managing this data and ensuring your project can meet its Program’s goals?  

You have a couple of options:

1. Spreadsheet

2. Database

Database a comprehensive collection of related data organized for convenient access, generally in a computer.

Spreadsheet a type of software that offers the user a visual display of a simulated multicolumn worksheet and the means of using it especially for financial plans and budgets.  

With a quick glance at a dictionary (dictionary.com if you were wondering) we can already see a deviation and how a database and spreadsheet differ and how one might be more suitable than the other in our use case.

3 Key Differences Between a Database and a Spreadsheet

1.     Data Formatting

In a Spreadsheet every cell is treated as a unique entity. It can store any type of information – a date, an integer value, a string name. Not only can we store different types of information and values in each cell, but we can also apply a specific format to these cells.

Databases generally work differently. They contain only raw data. Each cell is a container of a single data value, the smallest piece of information there is. The type of data contained in each cell must be pre-determined. This means that data ‘type’ errors are prevented – for example, in a field containing date values, should the user try to add a text string, the software will show an error and she will have the chance to correct herself. A spreadsheet such as excel will not prevent data type errors. Allowing the user to add data of any type to any cell

2.     Data Storage

In a spreadsheet, data can be stored in a cell, while in a database, data is stored in a record of a table, meaning you must count the records in a table to express how long the data table is, not the number of the cells. Cell fill, font colour, text colour, font – none of these things are relevant to a database. Storing the data in a manipulable format is all that matters.

Once we’ve got our data properly stored in a database format you can create stunning visualizations such as charts, maps, bars, graphs, and much more. Excel can give functionality to create charts, maps and graphs but options are limited and tied to the instance of the file containing the data

3.     How Calculations are Performed in a Database vs Spreadsheet

Another substantial difference is that, in a spreadsheet, different cells can contain calculations, such as functions and formulas. This means, that if you want to combine two integers, the result will be stored in another cell.

In a database, on the other hand, all calculations and operations are based on the existing data and are done after its retrieval. There is a specific feature, called “views”, similar to the tables, in which you can do a calculation. These objects also contain columns that can be normal columns like the ones in the tables or could contain a certain type of calculation. There is no way you can mistake a record of data with a calculation.  

What does it mean for your Semiconductor Fab when you choose a database over a spreadsheets

1.     Data Integrity

You can’t store different types of data in the same field – data type errors are inherently prevented in a database.

2.     Faster Manipulation of Data

In order to create graphs and output data, new tables and worksheets must be created. Instead of setting up spreadsheets or worksheets, in a database you can quickly create relationships between the tables, increasing the speed with which you could manipulate your dataset.

3.     Unlimited Volume of Data

Albeit powerful for many circumstances, spreadsheets have their limitations. Excel is incapable of handling over 1 million rows of data. This immediately induces us to look for a solution. Usually, the fix is to use databases, where having 2, 5, or 10 million records is not a problem. What might be more challenging is managing such vast swaths of data. But thankfully powerful integrations such as SQL + Tableau + Python allow us to handle even those.

4.     Better Multi-User Experience

Referring to the multi-user property, spreadsheets are lagging. We have all experienced the pain of the Excel that has been lost/corrupted. Or returning to Excel to discover it's out of sync with the latest design information because someone has created a copy and made changes independently of the master document.

5.     Data Consistency

A database unlike a spreadsheet is a stable structure, controlling access permissions and user restrictions. One person can make a change that is visible to everybody instantly. This feature increases efficiency and data consistency.

6.     Eliminating Duplicate Information

Considering data integrity and data consistency, using databases eliminates duplicate information, which is another way to save space and increase efficiency.