Robots in the Real World

OK then, let’s now take a look at some of the other interesting use cases with industrial and commercial robots.

Use Case: Security

Both Erik Schluntz and Travis Deyle have extensive backgrounds in the robotics industry, with stints at companies like Google and SpaceX. In 2016, they wanted to start their own venture but first spent considerable time trying to find a real-world application for the technology, which involved talking to numerous companies. Schluntz and Deyle found one common theme: the need for physical security of facilities. How could robots provide protection after 5 pm—without having to spend large amounts on security guards?

This resulted in the launch of Cobalt Robotics. The timing was spot-on because of the convergence of technologies like computer vision, machine learning, and, of course, the strides in robotics.

While using traditional security technology is effective—say with cameras and sensors—they are static and not necessarily good for real-time response. But with a robot, it’s possible to be much more proactive because of the mobility and the underlying intelligence.

However, people are still in the loop. Robots can then do what they are good at, such as 24/7 data processing and sensing, and people can focus on thinking critically and weighing the alternatives.

Besides its technology, Cobalt has been innovative with its business model, which it calls Robotics as a Service (RaaS). By charging a subscription, these devices are much more affordable for customers.

Use Case: Floor-Scrubbing Robots

We are likely to see some of the most interesting applications for robots in categories that are fairly mundane. Then again, these machines are really good at handling repetitive processes.

Take a look at Brain Corp, which was founded in 2009 by Dr. Eugene Izhikevich and Dr. Allen Gruber. They initially developed their technology for Qualcomm and DARPA. But Brain has since gone on to leverage machine learning and computer vision for self-driving robots. In all, the company has raised $125 million from investors like Qualcomm and SoftBank.

Brain’s flagship robot is Auto-C, which efficiently scrubs floors. Because of the AI system, called BrainOS (which is connected to the cloud), the machine is able to autonomously navigate complex environments. This is done by pressing a button, and then Auto-C quickly maps the route.

In late 2018, Brain struck an agreement with Walmart to roll out 1,500 Auto-C robots across hundreds of store locations.²² The company has also deployed robots at airports and malls.

But this is not the only robot in the works for Walmart. The company is also installing machines that can scan shelves to help with inventory management. With about 4,600 stores across the United States, robots will likely have a major impact on the retailer.²³

Use Case: Online Pharmacy

As a second-generation pharmacist, TJ Parker had first-hand experience with the frustrations people felt when managing their prescriptions. So he wondered: Might the solution be to create a digital pharmacy?

He was convinced that the answer was yes. But while he had a strong background in the industry, he needed a solid tech co-founder, which he found in Elliot Cohen, an MIT engineer. They would go on to create PillPack in 2013.

The focus was to reimagine the customer experience. By using an app or going to the PillPack web site, a user could easily sign up—such as to input insurance information, enter prescription needs, and schedule deliveries. When the user received the package, it would have detailed information about dose instructions and even images of each pill. Furthermore, each of the pills included labels and were presorted into containers.

To make all this a reality required a sophisticated technology infrastructure, called PharmacyOS. It also was based on a network of robots, which were located in an 80,000-square-foot warehouse. Through this, the system could efficiently sort and package the prescriptions. But the facility also had licensed pharmacists to manage the process and make sure everything was in compliance.

In June 2018, Amazon.com shelled out about $1 billion for PillPack. On the news, the shares of companies like CVS and Walgreens dropped on the fears that the e-commerce giant was preparing to make a big play for the healthcare market.

Use Case: Robot Scientists

Developing prescription drugs is enormously expensive. Based on research from the Tufts Center for the Study of Drug Development, the average comes to about $2.6 billion per approved compound.²⁴ In addition, it can easily take over a decade to get a new drug to market because of the onerous regulations.

But the use of sophisticated robots and deep learning could help. To see how, look at what researchers at the Universities of Aberystwyth and Cambridge have done. In 2009, they launched Adam, which was essentially a robot scientist that helped with the drug discovery process. Then a few years later, they launched Eve, which was the next-generation robot.

The system can come up with hypotheses and test them as well as run experiments. But the process is not just about brute-force calculations (the system can screen more than 10,000 compounds per day).²⁵ With deep learning, Eve is able to use intelligence to better identify those compounds with the most potential. For example, it was able to show that triclosan—a common element found in toothpaste to prevent the buildup of plaque—could be effective against parasite growth in malaria. This is especially important since the disease has been becoming more resistant to existing therapies.