How Flipkart is automating its supply chain with robotics, machine learning

Flipkart has 450 robots at its Bengaluru sortation facility
Tucked away in a nondescript lane on the outskirts of Bengaluru, is the sorting facility of e-commerce giant Flipkart. Walking into the hangar-like building, you can hear grinding metallic noises. That’s the sound of hundreds of robots, or cobots, as they are called, working seamlessly with humans.

The human workers place the products on the orange-coloured robots, and the machines sort them after scanning the encoded information on each parcel and placing them in the designated bins. “It is cool working with them,” says Nakula Punji, one of the thousand-odd employees at the facility.

Earlier, Punji would pick up products that customers had ordered, do the sorting, and walk back and forth for hours to get them ready for shipping to various destinations. With the robots to support him now, Punji is learning newer skills. Says Lalit Majhi, Punji’s co-worker. “I am now being trained in the safety aspects of working with equipment.” 

As it seeks to bring the next 200 million consumers online, Walmart-owned Flipkart, which competes with US rival Amazon, is betting big on robots as part of its technology-led initiatives. The company already sells products across over 80 categories, servicing all 20,000 pin codes in the country.

“If you have to scale business, the amount of space needed for sorting millions of packages, is humongous. Doing that manually is also prone to error. The idea behind using cobots is to work at scale and provide faster service to the customers,” said Pranav Saxena, head of robotics at Flipkart.

The company, which started using robots as a pilot early this year, has since quadrupled their number to 450 at its Soukya Road facility in Bengaluru. The robots, which are basically automated guided vehicles (AGVs), can sort 18,000 packages per hour. They are capable of working round-the-clock and they self-charge themselves at various charging points once their battery is drained. According to Flipkart, these AGVs or cobots can migrate easily, offering flexibility to deal with unplanned or seasonal demand spikes, such as during the company’s flagship festive sale, “Big Billion Days”.

“When you are growing fast, that scale also creates different kinds of problems, especially when you don’t know how much you will grow. For example, three years ago, you didn’t know whether people would buy more fashion or (electronics). To manage that, one needs flexibility and robotics brings that,” said Saxena.

Flipkart's sortation facility, which is as big as a football stadium, has two floors. A worker scans each package and puts it on a conveyor belt to take it to the specific floor. There, another operator feeds the shipments to robots, which take them to the nearby scanning machines to read the barcodes. Depending on the information, robots take the shipments to designated bins for items to be delivered to the specific pin codes. The robots use machine learning (ML) to identify the shortest route to a particular bin. 

The robots also communicate with each other in real time, whether it is to prevent a collision or for workload distribution. “We achieve 99.9 per cent precision with robots,” said Nikhil Vartak, a senior manager at Flipkart, and part of the company’s automation and research and development team.

All those handling the robots undergo rigorous training before they begin work on the shop floor. At the automation area, the employees are mandated to wear safety helmets, jackets and shoes. 

Experts see Flipkart’s use of robots as a stepping stone towards introducing automation across supply chain processes. “In India, companies have just begun to deploy robots at the warehouse level. Flipkart is leveraging its parent company Walmart’s acquisition of a robotics startup and using AGVs to increase sortation efficiency by almost 60 per cent,” said Ashwin Krishnamoorthy, associate vice president — technology and internet, at management consulting firm Praxis Global Alliance.

Warehousing and supply chain is one of the most crucial cost components of any e-commerce company. According to Ankur Pahwa, partner and national leader for e-commerce and consumer internet at EY India, now that these companies are shifting to specialised delivery timelines such as same-day and 4-hour deliveries, it has become even more critical for them to have effective logistics, warehousing and supply chains. “The biggest impact of using robotics in warehouses is that they increase the throughput by 5 to 10x. This technological upgrade in warehouses, combined with the upskilling of manpower, is necessary to substantially reduce logistic costs to cater to the massive growth in online shopping,” said Pahwa.

India, along with other global markets, is also testing upcoming technologies such as drones and self-driving vehicles to further reduce human intervention. Companies are also investing in ancillary technologies such as artificial intelligence and ML. AI in supply chains is helping in smart procurement, sorting of products and efficient demand forecasting, thus improving efficiencies as well as customer satisfaction.

“Amazon and Walmart have taken the lead in developing a drone workforce combined with RFID (radio-frequency identification) technology that could significantly automate inventory management. Drones can also help in real-time monitoring and resolve the out-of-stock problem,” said Pahwa of EY.

According to the International Federation of Robotics, the sale of industrial robots in India reached a new record, with 4,771 robots installed in 2018 alone. This is a 39 per cent increase over the previous year. India now ranks eleventh in the world in terms of annual installations of robots.

Though most of these robots are procured from other countries, Gurgaon-based GreyOrange is one of the earliest robotics start-ups in India that designs, manufactures and deploys advanced robotics systems. One of GreyOrange’s robotics systems accelerates the process of order fulfilment by queuing the items for picking. GreyOrange’s robots have been deployed by companies such as Mahindra, Delhivery, DTDC and Pepperfry.

Worldwide, the number of robots in use is 2.25 million — a three-fold rise over the past two decades, says a report by Oxford Economics. Trends suggest the global stock of robots will multiply even faster in the next 20 years, reaching 20 million by 2030.

Business Standard is now on Telegram.
For insightful reports and views on business, markets, politics and other issues, subscribe to our official Telegram channel