Hindalco adopts cloud-first strategy to modernise infra

Telemetry data from on-site machines have proven to be crucial, said Atanu Pramanic, Joint President and CIO of Hindalco Industries.

"Manufacturing is probably the last industry that is migrating from on-premise infrastructure to cloud," said Atanu Pramanic, Joint President & Chief Information Officer (CIO) of Hindalco Industries, at the NASSCOM Cloud Summit 2021 on Wednesday.

“In the past one and a half years, cloud migration has got a tremendous push,” Pramanic added, referring to the impact of the COVID-19 pandemic on the manufacturing industry.

Hindalco Industries, which generated Rs 1.3 lakh-crore in consolidated revenue in FY 2021, is one of the world’s largest aluminium rolling and recycling companies in the world.

It is an old economy manufacturer with a variety of machineries, as well as legacy applications. These applications were distributed, and ran on local servers until 2013, said Pramanic. “And then came the era of digitisation discussions becoming discussions of business agility.”

Large manufacturing enterprises typically start adopting newer technologies to reduce their time to market. And then, they begin migrating to the cloud from on-premise solutions.

Manufacturing companies now use telemetry to collect data from on-site machines using sensors or other devices, and automatically transfer the data to devices for monitoring. “Most of the manufacturing entity is a sticker of multiple versions, and multiple generations of machinery,” Pramanic said.

It is not always possible to get data from machines in the way metal-manufacturing enterprises can use. So, they have to invest in retrofitting these telemetry devices to collect that data in a usable form.

The pandemic impacted Hindalco's revenue and profit in FY 2020, as sales dropped 9.5 percent and profits plunging 31 percent to Rs 3,767 crore. The company is yet to return to pre-pandemic levels, and so technology has a huge role to play in ensuring efficiencies.

Hindalco is leveraging data to improve its production line in terms of resiliency, quality, productivity, asset utilisation, or condition-based monitoring. The quality and timeliness of data collection is critical.

“If you are not able to provide that data to the right person at the right time, then all this effort in connecting all the devices, putting in Internet of Things (devices and data) on the cloud, time to market, and all the stuff will fall flat,” Pramanic said.

However, manufacturing companies have mining locations where network connectivity is poor. So, for Hindalco, it cannot be “cloud-all or cloud-only” approach, which requires all applications to run on cloud. But it has a “cloud-first” strategy for the enterprise to shift its new applications to cloud.

“We have moved quite a substantial percentage of workloads to the cloud,” said Pramanic. “We have our own specific strategy to move a particular workload to the cloud.”

While deciding if an application should be on-premise or on cloud, Hindalco chose parameters that helped it evaluate application affinity to understand the effectiveness of the application. 

Ultimately, what matters is customer satisfaction. “The consumers of your application are the best people to judge where you should be,” he said.

Edited by Kunal Talgeri