This Indian IoT Product In China Shows How Factories Will Be Part Of The 4th Revolution
One of my fondest memories of growing up in a small town in India is making gashes in the bark of a tree on the way to school. The next day we would collect the gum which had seeped out. It was fun to use the natural gum to stick labels on our notebooks, even if it was messy.
Later I graduated to industrially produced adhesives, from simple glue to quick fixes for broken objects. But it’s only recently that I learned about the sophisticated technology and precision manufacturing involved in adhesives for such critical products as the wings of aircraft.
One of the leading manufacturers of adhesives of this kind is Henkel, headquartered in Dusseldorf, Germany. Its Dragon Plant (the name is a cultural homage to the plant’s sheer magnitude) in Shanghai, China, is the world’s largest adhesives factory, and aims to be the smartest one, too.
I was intrigued to discover how an IoT (internet of things) product made by an Indian startup, Flutura, in Bangalore is playing a transformative role in this. It’s part of a phenomenon some describe as the fourth industrial revolution.
Fitbit for machines
We’re fast getting used to a new world of smart homes, self-driving cars and virtual assistants. But an even bigger change is quietly brewing in our industries, thanks to huge advances in the collection and analysis of sensory data from industrial machines hooked up to the internet.
Just as feature phones had to give way to smartphones, the days of dumb industrial machines are also numbered. We are rapidly moving to an era of smart machines.
Derick Jose, co-founder of Flutura, describes its IoT product Cerebra as a “Fitbit for machines.” It captures data streaming from various sensors on machines to analyze their health and predict problems.
Think of the money saved when timely intervention, thanks to a trigger from Cerebra, prevents a breakdown in an oil rig. “One hour of downtime, interrupting the pumping out of oil, costs anywhere between $30,000 and $200,000 for one of our oil & gas clients in Houston,” elaborates Krishnan Raman, CEO and co-founder of Flutura.
Machines have always given out signals from sensors monitoring pressure, vibration and a host of other things. But once they’re connected to the internet, all those signals can be correlated and analyzed to make intelligent predictions and provide actionable insights.
Flutura has been working with OEMs [original equipment makers] like Stewart & Stevenson, Siemens, and JBT, which make oil rig pumps, energy equipment, and aerobridges, respectively. These are machines costing upwards of $1 million each, so any increase in efficiency or decrease in downtime leads to huge savings for their users.
In the case of Henkel, Bangalore’s Flutura is helping to transform the way quality is managed in making adhesives for clients like Bombardier, which uses a metal bonding glue for its aircraft wings. Here it’s going beyond smart machines to help create smart factories, because it extends across the whole manufacturing process.
Sandeep Sreekumar, who is strategizing Henkel’s smart factory move, explains to me that the IoT product from Flutura introduced in the Shanghai plant this year is part of a larger scheme of things that began in 2014. “IoT is not something you buy in the market and you suddenly become a smart factory. It doesn’t happen like that.”
Making an adhesive is a lot like cooking a dish. You have a set of ingredients and a method of cooking it. Various factors like quality of ingredients, quantity and timing determine how the dish turns out. Some things are hard to control, such as the influence of weather fluctuations on the characteristics of ingredients. That could affect the flavor of an onion or the viscosity of a glue material.
Here’s where a smart factory becomes a game-changer. What if you could predict how a dish will turn out instead of waiting to cook and taste it to decide if it’s good?
For Henkel, human errors, rework and quality testing account for much of the cost of making adhesives. A large inventory also has to be maintained in case of rejected batches. All those can be reduced drastically with a shift to prediction instead of inspection.
“We are creating an environment where you have a much smarter way to control quality,” says Sreekumar.
Sounds good in theory, but rolling it out on the ground involves a series of calibrated steps. Henkel began in 2014 with standardizing and simplifying the backend layers such as the ERP stack. The next year it began collecting data on manufacturing processes and digitizing the process flows.
This year it started connecting the machines and integrating the machine data with the process data to be able to derive predictive intelligence. That’s when a company like Flutura was brought in for actionable insights.
Cerebra gets into the act
Henkel makes hundreds of kinds of adhesives, each with its own mix of raw materials, and each material can be affected by changes in parameters like temperature and pressure. Flutura began to work with Henkel to build mathematical models that could handle this.
“The way we attacked it is in three steps,” Raman tells me. “First, Cerebra analyzed the quality to see what’s impacting it. Then it started predicting the quality. And the final step is for Cerebra to influence the quality in real time.”
Henkel’s dragon plant is all set to take that next step in its evolution to a smart factory. “The focus for next year is to pass information back in real time to the operator who is working, telling him to please do this instead of another step,” says Sreekumar.
The stakes are enormous. Factories will either have to get smart or get disrupted.
The scale-as-you-go model
For the Indian IoT startup, the deployment of its product in industrial player Henkel’s factories has fostered its expansion globally, from helping American firm JBT make smart airport equipment to a deal with Hitachi for the distribution of Cerebra in Japan.
It faces formidable competition from other industrial IoT products, such as those built on GE’s Predix platform. But the Indian startup, one of the first to enter this space, prides itself on its predictive algorithms and also gives its users options for customization and adoption lacking in bigger players. As Takashi Iizumi, president of Hitachi High-Tech Solutions, pointed out in a news release announcing its distribution agreement with Flutura: “The ‘scale-as-you-go’ pricing model from Flutura helps our customers to start small with valuable pilots and then scale the solution.”
This leads to faster adoption and implementation, which is becoming vital for equipment manufacturers as they race to adapt and get smart—or lose their way to other companies who will.