Rent a robot: robots as service providers springing up in a variety of industries
Posted by Upcoming Tech on
used with the permission of http://thenetwork.cisco.com, by Anne Field
Robotic systems can offer businesses a great many benefits: reducing costs, increasing productivity and freeing up humans to do less repetitive, more interesting tasks. But for many small to medium-sized businesses, buying, installing and maintaining them can also be quite pricey, especially if it means throwing out and revamping existing production systems.
That’s where “robots as a service”(RaaS) comes into play. Recently, a growing number of companies have started offering customers the option of leasing or renting robotic systems either by the hour or with a monthly subscription, instead of purchasing and installing them on their own.
The advantages are similar to outsourcing, according to Frank Tobe, editor of The Robot Report. On the one hand, customers don’t have to bear the cost of installing robots. On the other, services providers can invest in and keep up with the latest technologies, while spreading their expenses out over many users, targeting industries ranging from agriculture to material handling. “They’re finding that economies of scale benefit the service provider,” says Tobe.
Here are three sectors where RaaS is taking off.
Robots as a Security Guard
Security—that is, ensuring office buildings, warehouses and other locations are safe—is one area that has attracted RaaS players. These companies provide robots that replicate human security guards, but allow customers, in effect, to pay as they go. “To get many businesses to buy a robot for $50,000 to be security guards is like pulling teeth,” says Tobe. “But a service that’s maybe $7 an hour, particularly if they can stop at any time, they’re happy to do it.”
Knightscope is one such provider. The three-year-old Mountain View, California, company has a subscription-based service, with security robots for everything from corporate campuses and shopping centers, to data centers, airports and sports stadiums. (The company’s term is “machines as a service”).
Called Autonomous Data Machines (ADMs), the gizmos come in two sizes—a five-foot-tall, three-foot-wide, 300 pounder and a lighter version that’s four feet by two feet. Both can be programmed to operate in a particular geographic area and taught to recognize what is and isn’t supposed to be there. Using sensors, lasers, 360-degree video and two-way audio, they roam their prescribed area, detecting anomalies, like a car parked for too long in a certain location or a person wandering around a facility after hours. With a clear line of site, one robot covers about half a million square feet, according to co-founder Stacy Dean Stephens.
Humans do get involved. Robots send the data they collect in real-time through the cloud to a centralized data center or to an individual with a smart phone. They then make decisions about next steps and they can replay events later on, to look for patterns and other insights.
Security robots may be particularly handy for hard-to-fill spots like the late-night third shift, as well as times when a customer doesn’t usually have a guard on premises. Jeff Debrosse, co-founder of San Diego-based security technology company NXT Robotics, points to one university considering using his service for patrolling around the administration area at night, when there are no guards. “They’re filling a hole,” he says.
It’s not something consumers usually think about when they’re ordering a product online, but a warehouse worker has to find that item on the right shelf, grab (or “pick”) and scan it, put it in a carrier and transport it to the appropriate place for shipping. It’s a time-consuming, mind-numbing, error-prone task that’s growing in importance as e-commerce—and expectations for fast delivery—become the norm.
Robotic systems, of course, can make the process much faster and more efficient. But there are thousands of warehouses in operation, many of which lack the resources to install their own systems.
With that in mind, Agoura, California based inVia Robotics was founded in 2015, to provide a robotic picking and carrying system. According to co-founder Lior Elazary, the typical customer pays about $10,000-$20,000 for setup, and about 10 cents a pick. A robotic system that handles 50,000 items a month would cost the company $5,000 a month. That’s compared to the $2 million to $3 million cost of putting in an entire system from scratch. Customers also can start with a small number of robots and expand as demand increases.
How does it work? A robotic management system is installed in a warehouse that manages both robots and people, assigning tasks and making sure everyone—human and electronic—knows where to go and does so in the most efficient manner possible (That can be done remotely from another area). Once the system is up and running, the robot traverses the floor to map out the location of items. If, later on, it doesn’t recognize something, it displays an image on a screen and asks what the item is; a person who is supervising, probably at a different operations center, can then label and identify the image.
When they have their marching orders, robots can grasp objects ranging from jewelry to cases of soda, and then hand the object to a carrier device, which receives and delivers the material to another area for packing and shipping. The system also counts how many items are left and verifies that the system picked the correct object.
Of particular interest are drone –based systems. They can fly over fields, taking pictures of crops to see where, for example, a farmer may need to increase the amount of fertilizer or decrease the level of pesticide. SenseFly, for one, a subsidiary of Parrot Group, has what the company calls a “drone to tractor” system. Customers can lease it from the Lausanne, Switzerland, business, buy it outright, or pay another company to provide the services. The robotic device follows a programed flight path, capturing images which are then transmitted to computers or a smartphone. Those pictures are used to create a more detailed map of the crop, along with a customized recipe for a plan of action—say, how many pounds of fertilizer should be used in a particular area, or even the number of seeds that should be planted. Then the map is exported directly on a tractor’s console.
Not every industry is receptive to RaaS, of course. NXT Robotics’ Debrosse originally targeted the hospitality industry, offering robots for such tasks as delivering luggage to guests’ rooms. But potential customers showed little interest. After seven months or so, he added new sensors and focused on security. “Once we did that pivot, demand opened up,” he says.