There has been a lot of talk about the enormous potential and market opportunity for smart factories. Yet, for all the talk, the reality on the ground for a lot of enterprises in manufacturing is that their machines are not yet connected to any network. Before they can even talk about digitization and predictive maintenance, they need to first get connected.
There are many sectors in which this lack of connection has traditionally been a pain point (consumer packaged goods, for example), but other sectors, particularly oil, gas, and automotive, have had the ability to measure and monitor things through connectivity and sensors for a long time.
However, to date they have been using data and analytics primarily for engineering value and manufacturing KPIs. Enterprises need to start looking at data from the angle of its business impact. For instance, how can data be utilized to minimize downtime? Analytics have the potential to totally reconfigure the factory, but they must be utilized across the enterprise, from labor deployment to supply chain management.
Areas of Opportunity
Though each industry can utilize the interconnected nature of Industry 4.0, certain sectors will especially benefit. Companies in industries with products that rely on optimal mixtures of materials or chemicals are especially ripe for innovation.
For instance, there is a massive opportunity to build and leverage algorithms that can create recipes for chemicals by looking at the composition of a recipe and coming up with the optimal mix. This algorithmic blending brings value by quickly discovering relationships between chemicals and bringing new products to market. It saves valuable time and effort, as this process, when done manually, is very laborious. Humans would have to analyze each individual chemical and the relationships between them in a painfully time-consuming process otherwise.
Machines trained on chemical data not only do this much faster; they can also present the information in a visual manner that maximizes further analysis and immediately demonstrates the new product’s value. Armed with algorithmic blending capabilities, R&D teams can suddenly generate a series of new products with massive revenue potential. Moreover, algorithmic blending can be used for any chemical product, creating an enormous spectrum of applications.
In pharmaceuticals, for example, drug synthesis could be revolutionized through algorithmic blending. Companies have the internal data on their chemical compounds, but the data is not organized, and so algorithms cannot be trained on this data to optimize mixtures. The same goes for oil and gas, where they are mixing products such as petroleum, chemicals, and lubricants. With countless applications, algorithmic blending has the potential to revolutionize some of the largest and most profitable sectors in the world. It will be a key feature of Industry 4.0 in the future.
How Can Data Help?
It is easy to build solutions that leverage the power of analytics, but if they are not used, then the solutions are a waste. How can enterprises build solutions that work for them and drive value?
First, enterprises should start by creating a small pool of internal capabilities and people who understand how analytics can actually solve domain-specific problems. A huge limitation for companies is that only some have an internal team to focus on analytics initiatives. Forming cross-functional teams with one layer of people internally supported by another layer of external partners is a good place to start. Without internal analytics capabilities and buy-in from the top, analytics solutions become the shiny new toy for one to two quarters and then are forgotten.
Second, in order for AI and predictive analytics solutions to interface well with existing platforms, manufacturing should be able to integrate the cloud with manufacturing ERP solutions. These solutions work as an integration layer across all elements of an operation, vs. as a one-off solution.
Finally, it is important to implement pilot programs. These pilot programs should be designed to be repeatable. The big question enterprises should ask about ROI is how quickly they can repeat the program across other divisions in the enterprise. The goal is to be able to build one pilot program and then deploy it multiple times across the company and across divisions.
From Steam to Dreams
Built on the back of the steam engine, the First Industrial Revolution fundamentally transformed the way people and organizations worked. Today we are witnessing how data is transforming organizations in a similar way with Industry 4.0. The factories of tomorrow will be totally connected, allowing us to leverage the full impact of data analytics.
Venkat Viswanathan is Founder and Chairman of LatentView Analytics
