Retail Is the Prime Testing Ground for the Next Wave of Mobile Robotics

With the conclusion of the Mobile Robot Summit in Denmark, it is an excellent time to take stock of the most dynamic part of the automation industry: the popularization of autonomous vehicles, mainly for material handling within the supply chain, has bypassed the experimental stage and is being implemented on an increasingly grand scale. Dramatic improvements in accuracy and efficiency in sensor fusion solutions, alongside advancements in Simultaneous Localization and Mapping (SLAM), have made autonomous mobility readily available for the vast majority of form factors.

While much attention has been focused on the promise of self-driving cars, the enormous venture funding allocated to their engineering teams has, if anything, not lived up to the timeline envisioned. Nevertheless, the self-driving car hype is a slow burner, with large automakers and technology companies making long-term bets in the hopes that autonomous driving will open up new opportunities in last-mile delivery, Mobility-as-a-Service (MaaS), and other opportunities. From a robotics standpoint, mobile autonomy is still the most consequential area, as can be highlighted by frontrunner Waymo’s hiring of 13 employees from the now-defunct consumer robotics startup Anki.

But while Chart 1 highlights the discrepancy between road-based robotics startups and other mobile robotics companies as far as funding enthusiasm goes, it is the deployment of robots in indoor environments for material handling that is taking off, as shown in Chart 2.

This figure does not provide a comprehensive display of mobile robotics companies, but instead focuses on a select number of the most well-known startups and advanced Autonomous Mobile Robot (AMR) developers. The majority of these companies are offering mobile platforms for A to B delivery, with Brain Corp being the main entrant deploying robots for floor cleaning in the retail space. The company currently has 3,000 autonomous floor cleaners deployed, utilizing vehicles from major Original Equipment Manufacturers (OEMs) like Tennant and Nilfisk. BlueBotics is one of the more interesting material handling providers. Like Brain Corp, it is an Autonomy Solution Provider (ASP) that develops vehicle-agnostic navigation solutions to be applied to a range of form factors. This includes cleaning robots, material handling vehicles, forklifts, and heavy industrial vehicles, with its largest automated vehicle being a platform designed to carry 20-ton payloads. In addition to automating vehicles, BlueBotics offers consultancy to companies on developing robots from the ground up, and views this as a growing part of its consultancy services, with customers including ABB, Nestle, and the European Space Agency (ESA). The idea of consulting with OEMs to build the next generation of mobile robots from the ground up in addition to offering navigation represents a step forward for ASPs that is further replicated by Brain Corp’s involvement in powering SoftBank Robotics’ smaller floor cleaner, Whizz. ASPs’ success rests partially on their ability to develop quick solutions in existing vehicles with decades of development behind them, which allows for quicker design and deployment cycles.

With a handful of AMR developers deploying thousands of robots, this industry is guaranteed to grow extensively in intralogistics and warehousing and has been deployed succesfully in healthcare through companies like Aethon. However, significant challenges remain.

The value proposition for mobile robots in intralogistics is simple: stuff needs moving, and drivers bring with them cost and safety concerns. What is more, the moving of materials is a potentially wasteful part of the working process that has to be optimized. Outside of these established markets, however, the value of mobile robots is more varied.

Retail is the perfect case in point: there are robots for floor cleaning, data-collecting systems for inventory, automated tugs for material handling in the back room and on the shop floor, and even robots to provide security. This is all happening in locations that often have elevators and connected infrastructure in addition to being open to the general public. What is yet to have happened is effective transitions between indoor and outdoor environments. This, as well as linking up Brick and Mortar (B&M) retail with last-mile delivery, could become a key competency for certain retail configurations. Companies are increasingly looking for ways to bridge this gap. Amazon Mobile Robotics, formerly Canvas Technology, explicitly advertised this as a competitive advantage based on its vision-based navigation. Technology developer Humatics has also focused on facilitating seamless navigation from indoor to outdoor environments through ultra-wide band beacons that allow for localization accuracy down to millimeters, with MIR and Vecna being notable partners.

Big retailers have all of the requisites to make scaling of robotic products much easier than, say, hotels do. Kroger has developed a partnership with pioneering last-mile delivery company Nuro and Walmart is leading the industry in technology development, albeit under the radar. Surprisingly, Walmart has filed more patents for drones than Amazon, the poster child for robotic deployment at scale.

Walmart’s most high-profile partnership to date is with Bossa Nova, one of the best-known robotic startups. So far in 2019, the robotic company has deployed 350 systems for inventory management across the retailing giant’s stores, with roughly one robot required for each store. Though Bossa Nova is the most advanced in deploying mobile platforms for inventory, there are a host of other developers targeting the space including Badger Technologies and Simbe Robotics, the latter of which recently received US$26 million in funding in September 2019.

The value of retail as a testing ground for the next wave of robots is not lost on adjacent technology providers. AT&T has partnered with Badger Technologies to test 5G-enabled multi-access edge computing to help vendors and end users cope with the increased bandwidth requirements deploying robots incurs. Further opportunities for adjacent technologies lay in fleet management, management software, and Vehicle-to-Vehicle (V2V) communication.

Looking at the numbers, robots in retail is booming. Brain Corp has deployed 3,000 industrial floor cleaners across retail environments, Bossa Nova has deployed 350 inventory robots in Walmart, and Simbe has deployed in retailers Decathlon and Schnuck, expecting to deploy 500 systems over the next year. Bossa Nova expects shipments to double every year for the next five years and expects to scale up to 10,000 deployments in that time.

Still, significant challenges to deploying robots in retail remain. Bossa Nova notes how it collects up to 10 terabytes of data today and cannot send this amount to cloud services. Such strains will only grow with scaling. Telecommunication companies and cloud service providers will be vital to large-scale deployment of mixed fleets in the future. This service model will likely be preferred due to the challenges retailers would face deploying the technology themselves.

So, what is the biggest use case opportunity? At present, Brain Corp is very far ahead in deploying systems at scale, while new opportunities in material handling from small robots and tugs, like those from Odense startup Coalescent Mobile Robotics (CM-Robotics), could become more valuable as fleet management becomes more attainable.

Thinking more strategically, B&M retailers know they have a future in the supply chain. Amazon itself, the e-commerce slayer of hundreds of (B&M) businesses, acknowledged their intrinsic value when it bought Whole Foods. However, in order to compete, these retailers have to innovate and change. Part of this change is the shift to an omnichannel approach that acknowledges the need for multiple channels to meet the varying demands of the customer, who is becoming increasingly demanding. A strong force behind the robot-related overtures of Walmart and Kroeger is that last mile delivery is the great oppurtunity for cost reductions in the supply chain, and 90% of U.S. households. Due to this, retailers want to become distribution centers in their own right, so as to compete with e-commerce providers. A problem with this is that while inventory accuracy for e-commerce providers is 99%, the accuracy for brick and mortar retailers is closer to 60%, in part due to legacy infrastructure. This is the real value proposition of companies like Simbe and Bossa Nova: to become data companies that optimize inventory management and help retail stores radically extend their value proposition to the customer.

This is the biggest long-term robot-enabled shift in retail. Improved floor cleaning, material handling, and security through robotic platforms represent excellent additional opportunities that will feed into the improved data mosaic. Those robotic companies that can focus on data-related services stand the best chance of creating common platforms and scaling to the requisite size necessary to meet the demands of retailing giants like Kroeger and Walmart. The automation is important, but the data is key.