I am convinced by the power of data, machine learning and AI at every level of a company. We will build new tools, connect them together, and make them smarter and smarter. The change will come slowly. So slowly that you actually may be the one who makes it happen. In the following, I’ll try to describe my opinion of the actual state of the traditional industry and try to propose a way to start to inject some data in it!

1. Context

Data is the key value of the 21st century. In the last decade, our capacity to register and analyse data has increased along with the improvements in hardware parts and decrease in parts’s prices. It became really easy to set up your own database or data recording system, to define your KPIs and to start building monitoring system. Knowledge has also spread at the speed of light with Internet, and every one is now able to develop a data tool with the help of forums and educational platforms.

While the web industry has already come a long way, manufacturing industry will have to follow the same path. Small or big, every company will have to make the steps into the data world. In most developed countries, government and universities have already understood the value of data and are pushing strongly to create a real culture around it.

Sensors have become more reliable and cheap, they are easy to deploy on your different machines or in your production environment. In several industries, engineer will record data along the process and use it for their own purposes.

2. Diversity of application

Why is data valuable for your company? Well, there are several layers to how we can use data. Data can concern everyone in a company : supply chain, engineering, quality assurance, IT, production , top management. Everyone has a different need for in data and would like to observe different phenomena.

Different goals and different approaches must be developed. While top management will be more interested in BI applications and consolidated data in order to make important decisions about the business, engineers may instead want to develop products which includes machine learning tasks and models. The quality assurance department may be interested in developing a defected products recording system and trying to link it with all the parameters recorded along the process to try finding a solution to their issues, whereas the maintenance department may have set an intelligent intervention tracking system which can anticipate the future breakdown of machines (predictive maintenance). A marketing department may be interested in finding new clients or focusing on their current customers sentiment analysis. Human resources will use data to benchmark salaries or predict departures of their beloved employees. Data has marketing value as well as strategic value, and the only way to reveal its true value is by building algorithms around it. This is when the true potential of the data will reveal itself.

	Section	Examples of application
	Quality	Defect recording system Rootcause analysis Monitoring Automatic defect classification
	Management	BI Monitoring Dashboards
	Maintenance	Maintenance history Preventive maintenance Spare parts management
	Production	WIP Lead time optimization Monitoring
	Engineering	Product development Machine learning application
	Marketing	Trends of the market Customer's sentiment analysis
	Human Resources	Salary Benchmark Employee resignation anticipation

Most of the companies have achieved the first step toward data-driven environments. Most of its made the transition to the digital world and have begun to record data for some of their activities. Nevertheless, the value and knowledge we can extract from data is under exploited.

3. The 21st century company

Fear of losing their time, fear of not being able to finish the project, managers may feel timorous to launch their data revolution for some reasons. Hopefully, with good guidance and a clear vision of the company process, the risk of failure in your data project is practically null. It is guaranteed you will be able to extract value from your data.

With this known, we have to keep in mind that it will not be easy; we will have to face many kinds of challenges during a data project. It takes a lot of time to get specialized in each field of data. To represent the diversity of domains, here are few activities that we can have associated with data, each having its own tool and knowledge associated : big data, machine learning, data visualization (communication), data management, data cleaning, user interface building, … you need the skills of a data scientist to reveal the deepest information hidden in it.

Steps :

• Identify your key value data. Depending on your project / specialty, the first step for everyone is to define a clear vision of where your business is at and which key values you will need to record. Define your sources of data (internal : databases, files, pictures, etc…), or external (social networks, open databases, API, etc…) and build the tool to retrieve it all properly.
• Start recording data, make it consistent, secured and accessible to the right people. Segment your data system.
• Start developing your ideas. Gather teams around projects (IT, specialists, data scientist) and start building your new products around data.
• Continue to develop your skills and become a key company in the 21st century.

The 21st century company will have to face the data challenges if it wants to survive in the upcoming competitive market. An IT department should not take for granted the value of data. Their task is to provide and secure the infrastructure to manage the data and the tools relevant to the data. Data scientists and the IT department will then have to work hand in hand in order to become a data-driven company.

Data can give you great insights into how your company works today, but with some advanced tools it will also be able to predict the future and help you to make the right decision at the right time.

I hope I could convince you that you should go deeper into the data world and start your data revolution. I am also passionate about this kind of topic. If you need help to start the journey to the 4.0 version of your company, please do not hesitate to contact me. I’ll be glad to participate in such an adventure. Lots of challenges lay ahead, but at the end of the journey, you will have made the most important step towards the 21st century factory.

Let’s unite data scientists and experts to develop the next generation of industries!