Robotics-as-a-Service: Transforming the Future of Retail
By Dr. Swagat Kumar, Head of Robotics Research Group, TCS and Chander Shekher Joshi, Innovation Evangelist, TCS
It is past 3 am at a fulfillment center in Southern California. Brad has been working the night shift. As a picker, Brad generally covers three miles an hour, works long hours to meet the holiday season rush, takes only minimal breaks, and yet shows no signs of fatigue. These tireless efforts by someone who is just one foot tall seem phenomenal. If this seems like a scenario from a Christmas fantasy about Santa and his elves, think again. There are several others like Brad— robots who work behind the scenes at retail warehouses and distribution centers, helping deliver merchandise to customers every day.
Robots have today moved beyond the realm of traditional jobs such as assembling of parts in the automobile and manufacturing industries. From military to healthcare to retail, robots are slowly becoming ubiquitous, delivering newer efficiencies. When combined with the digital forces such as cloud and analytics, these smart companions can have a positive impact across the retail value chain.
Robots in retail stores — fantasy or reality?
The digital economy is accelerating the pace at which both retail enterprises and consumers adapt to exponential advancements in technology. Smart robots are already triggering retail innovation. For instance, if you hold a random screw in front of OSHbot, the multilingual robot at Lowe’s Orchard Supply Hardware, it will scan the item and lead you to the aisle that has it. At Hointer, a Seattle-based apparel store, robots deliver selected clothes directly to the fitting rooms.
In the next 5-10 years, when customers walk into a store, they are likely to be greeted by a robotic assistant. Robots in future stores will be able to understand the store layout and real-time product availability as well as interpret gestures, interact with customers, converse in the local language, and make product recommendations.
The impact of robots on the retail value chain
Robots are likely to play an important role throughout the retail value chain including both front-end and back-end operations. Here are a few possible use cases in the areas of customer assistance, employee assistance, and compliance audits.
Deploying robotic carts: Navigating through aisles searching for the right product has always been a customer pain point. A robotic shopping cart can assist shoppers by guiding them to the right aisle. It can also perform an auto-checkout as shoppers add items to the cart or cancel the checkout when items are removed. As an extension of this idea, customers would be able to capture, store, and share images of products from a virtual shopping cart. Based on the shopper’s purchases across different stores, the cart maps the path to the commonly purchased items, expediting the purchase cycle.
Delivering personalized engagement: In a fashion store, robots can serve as shopping assistants by inferring personal likes and dislikes from non-verbal cues such as facial expressions, gestures, and body language of customers. This information can be augmented by scanning customers’ physical attributes such as height, hair and eye color, and skin tone. The multilingual robot can then recommend off-the-shelf items based on the customer’s budget, previous purchase history, and personal characteristics.
Offering value-added service: Large retailers typically have separate counters for ‘in- demand services’. Robots can be used to provide a range of value-added services such as currency exchange, gift-wrapping, price comparisons, and dispensing loyalty coupons. With robots taking over monotonous tasks, store personnel are free to focus on servicing customers at the checkout counters.
Supporting faster fulfillment: Robots can be deployed at stores for quicker fulfillment of online orders, especially during peak hours and the holiday season. Store personnel spend a lot of time making trips to the store inventory, identifying items that are ready for pickup, loading them into the cart, and delivering them at the pickup counter. A robotic assistant can take on these routine tasks to ensure faster fulfillment.
Routine checks and compliance audits
Minimizing out-of-shelf scenarios: At least 25 per cent of ‘out of stock’ items are in the store—just not on the right shelf. Robots can be deployed to move around the store constantly, capture images of aisles and products, and send alerts to a central system for misplaced items, empty shelves, and low stock situations. Robots can be used to automatically replenish shelves.
Ensuring planogram compliance: Despite the use of sophisticated automated systems, consumer packaged goods (CPG) companies incur large planogram compliance costs, both in terms of money and time. Robots can be deployed for out of shelf (OOS) checks and planogram compliance, thus helping CPG companies save thousands of dollars. Robots can capture images of aisles and products, which can then be shared with CPG companies to gather insights on product performance and design store-specific planograms.
Mitigating shrinkage and theft: In the U.S. alone, retailers lose an average of 1.3% or $54 billion of sales through shrink per year. Surveillance robots deployed at stores can help identify suspicious activity based on motion and blinking eye patterns of shoplifters and store personnel. They can also capture images of trespassers and send alerts to the store manager to help reduce pilferage.
Understanding the combined power of Robotics, Analytics, and Cloud (RAC)
Leveraging autonomous, multipurpose, and bespoke robots is likely to have lasting impact on the retail value chain. However, winning in the digital world will require harnessing the combined power of the digital five forces—Big Data and Analytics, Mobility and Pervasive Computing, Cloud Computing, Social Media, and Artificial Intelligence (AI) and Robotics. Let’s examine how the merger of Robotics, Analytics, and Cloud (RAC) has the potential to fundamentally change the way retail enterprises conduct their business and engage with customers.
Consider a scenario where a fleet of robots designed to perform multiple front-end and back-end functions are deployed across different stores. The robots capture critical customer-specific and store related data and insights such as customer preferences and inventory turnover by geography. The data is stored on a centralized cloud system that can be retrieved by different stakeholders to make informed decisions. For instance, CPG companies can better understand how products are placed in the store and the corresponding impact on sales. With the availability of geography-specific customer browsing data, they can also come up with location-specific planograms. Advanced analytics helps CPG and retailers to consolidate data from all stores to better understand customer purchasing behavior and track real time demands. As shown in Figure 1, such insights can help distributors plan ahead effectively. Analytics can also be applied to other retail functions such as supply chain management, marketing, and business operations.
Leveraging RAC for integrating customer browsing and purchasing behavior with the supply chain
Figure 1: Leveraging RAC for integrating customer browsing and purchasing behavior with the supply chain
The research or ‘Explore’ team works on advancing state-of-the-art in technologies related to customer domains. The engineering or ‘Enable’ team picks the most promising ideas and creates applications. Ideas with the highest market traction are scaled by the ‘Exploit’ team and adopted by the mainstream business as customer offerings. Many businesses may have these innovation processes in some form or another. But we go a step further with the ‘Evangelize’ team that consists of people who understand technology and connect with business units and customers about relevant work in our Innovation Labs. As they are in constant touch with customers, this team understands customer and market expectations and provides feedback to researchers. In the past, we have had to modify our projects because customer expectations were quite different from the original research motivation.
The emergence of a new business model: Robotics-As-A-Service
Deploying a fleet of robots, maintaining them, and integrating robotics with multiple databases across the retail value chain is challenging both in terms of budget and effort. When procured as a product, the robot is bundled with the software and hardware capabilities needed for executing specialized tasks autonomously. Managing hundreds of robots spread across geographical locations adds to the challenge. A new business model where robots are offered as a service rather than as a product can help address these challenges and concerns. The advantages of the Robotics-As-A-Service (RaaS) business model based on an RAC framework are outlined below.
Minimal upfront costs: Since robots are programmed to execute complicated computation on a cloud server, the costs associated with employee adoption of an enterprise RAC platform are minimal.
Smarter networks: Robots are likely to be designed to be part of an intelligent ambient system. This means that they can interact with surveillance cameras and RFID antennae, enabling a smarter business network. For instance, robots can analyze customer footfall captured from surveillance cameras and send alerts to store personnel on shelf replenishment, thereby eliminating out-of-stock situations.
Economies of scale: The RAC service provider can deploy a fleet of robots to support multiple business functions, reducing the average cost of ownership. Technologies such as cloud that enable resource sharing can help further reduce costs. In addition, with only a single human operator needed to manage, monitor, and control multiple robots, companies can eliminate several overheads and associated costs.
Easy upgradation of skills: For a robot, learning a new skill can be as simple as downloading a mobile app. This can help reduce the effort and time taken to learn new skills, empowering robots to improve performance quickly. For instance, business solutions developed by robotics operators can be recorded as a log that can later be retrieved by a robot for solving similar problems. Robots can not only learn from human operators but also from other robots.
Challenges versus benefits: Weighing the pros and cons of RaaS
The proposed business model—leveraging robots as a service—comes with its fair share of challenges. One of the biggest challenges lies in integrating the three technologies of RAC and collaborating with different stakeholders to set up this model. Working with robots can impact employee morale as employees may perceive them as a threat to their livelihood. As with any new technology, customers may also be apprehensive about interacting with robots in a store.
Despite these challenges, the RaaS business model may be an effective alternative for retailers contending with escalating training costs per employee or cost per square foot.
Is RaaS only for big box retailers?
The RaaS business model can be leveraged by both mid-segment and big box retailers, with each opting for one or more services based on their requirements. For instance, a medium-sized retail store may prefer using a simple robot for basic monitoring and stock assessment on shelves. On the other hand, a big box retailer may opt for services that offer special capabilities such as interacting with customers through speech and vision. While the physical infrastructure cost would increase linearly with expansion, the operational cost will reduce exponentially as the same software can be used for managing several hundred robots.
The RAC service provider can be responsible for maintaining these robots and providing analytics based on the data gathered across multiple stores and brands to CPG companies (refer Figure 2). This will provide visibility to warehouses on upcoming demand and to customers on product availability. Again, this is something that both mid-segment and big box retailers can benefit from.
The RaaS business model
Figure 2: The RaaS business model
Big box retailers such as Staples, Walgreens, Gap, and The Office Depot, and online players such as Amazon are using robots in their distribution centers. These robots fetch merchandise and deliver it to workers for packaging and shipment. Amazon has installed 15,000 robots across 10 warehouses in the U.S. to get packages out more quickly. This holiday season, robots have allowed Amazon to ship packages in as little as 13 minutes from pick stations. This is a dramatic improvement over the average hour and a half taken to ship packages at older centers. It isn’t just back-end operations that are benefitting from the use of robots. From customer interactions to store operations, robots are delivering efficiencies and improvements on several fronts.
It is clear that retailers cannot ignore the rise of robotics and the potential benefits of RAC in transforming the future of business. To get started, retailers need to identify jobs that are manual, mundane, repetitive, or hazardous. Deploying robots for such types of tasks will allow retailers to improve productivity and enable employees to focus on high value-adding tasks such as servicing customers. Retailers will also need to protect employee morale by using preparatory and counseling programs that position robots not as replacements but as a way to augment the organizational workforce.