Robotics-as-a-Service is becoming increasingly common in warehouses. Thanks to robotics' growing autonomy and affordability, robots are anticipated to become the rule rather than the exception in those contexts by 2025, according to ABI Research.

That prediction is corroborated by Robotics and the Future of Production and Work: “With integration of artificial intelligence (AI) and other improvements in robotics (e.g., better machine vision, better sensors, etc.), robotics promises to see significantly improved pricing and performance over the next decade.” (Read 7 Women Leaders in AI, Machine Learning and Robotics.)

A New Automated Warehousing Solution

It’s not just advances in technology that are making highly capable robots affordable, but a new business model that puts them in reach of more businesses than ever before. The ABI Research article explains: “Robotics-as-a-Service models mean that large CapEx costs can be replaced with more accessible OpEx costs that are directly proportional to the consumption of technologies or services, improving the affordability of robotics systems among the mid-market, further driving adoption.”

A major player in the RaaS space is InVia Robotics. InVia Robotics’ Co-Founder & Chief Technology Officer, Rand Voorhies spoke about how RaaS uses the Internet of Things (IoT), AI, and other advanced technology to deliver economical and efficient solutions to businesses. (Read 5 Defining Qualities of Robots.)

Applying automation to retrieval from warehouses itself is not new, but the agility and rapid deployment made possible by the robotics made available as a service is certainly a game changer for businesses.

Voorhies said that how “automatic storage and retrieval systems” worked in the past was “like a vending machine 20 stories tall.” Such a structure would take 6 to 12 months to build and cost millions of dollars. Any change in needs for the setup would entail another hefty layout and substantial lead time to prepare a new warehouse.

Robots Built for Warehouses

In contrast, the initial setup and deployment of their RaaS can be completed over a single weekend. The robots that arrive are all ready to plug in and start work. They all connect to centralized server either in warehouse or cloud, depending on customer’s internet connection. They can then grab things by themselves and drive by themselves, as they pick up on the direction rather “like air traffic controls with path authorization,” he said. (Read The Impact IoT is Having on Different Industries.)

Voorhies explained that their robots are designed specifically for their warehouse functions, to navigate the aisles, find what needs to be retrieved, and safely interact with humans and each other. The first prototype was built in the CEO’s garage, but, as Voories explained, it couldn’t come into its full functionality “in isolation.” They had to learn how to adapt it in a real warehouse where they out to “get it running and figure out everything about logistics.”

The company founders weren’t working in isolation either. They spent "a lot of time touring a lot of warehouses and speaking to a lot of warehouse managers," he said. After getting the basic setup down, they worked on motion studies to optimize the robot’s efficiency and the user interface. They have a lab that simulates the type of pick station robots would work on in warehouse, including a pick station, order containers, and a user interface.

The people in the warehouse are spared walking around to retrieve the items needed to fill an order. Instead the robots bring them the boxes that contain the item, and that the human gets the directions of how many to take out and where to put them. Robots bring over boxes, and a person scans the item put in an order container. The robot detects the item was scanned and then can move on with the container to either deliver it to another station or put it back in its place.

That way, each retrieval can bring an item needed for multiple orders to fill them faster. Voorhies offered the example of a robot bringing in a box of socks from which 39 pairs will be taken to fill 25 different orders. Another box would be brought with hats to add to 15 of those same orders, and so on.

You can the robots in action in the video below:

Traffic control

The intent is to make the retrieval of items as seamless and efficient as possible and to allow businesses to meet their own customer demand for deliveries. “You can think of it as a magical, perfect conveyor belt,” Voorhies said.

But that perfect transition requires a lot of programming and AI capability to get them where they need to go as quickly as possible without hitting anyone or anything along the way. That’s why the “traffic routing system” that has to manage the route of hundreds of robots within the fairly tight space of a warehouse.

It draws on “really smart optimization algorithms” that “recalculate about 10 times a second,” as well as some onboard proximity sensors. As in IoT in general, it’s all about pulling together information in real time and directing action accordingly.

Focus on service

As Voorhies explained, "the customers don’t buy the robots but the service, the order fulfillment." His company delivers a “full stack solution.” That means that: “Orders come in from the internet into our system and then trickle down in inventory management and into our control of robots.” The robots are not controlled individually but rather, “work as a cohesive swarm together.”

Accordingly, it’s in the interest of both inVia and the customers it serves to get just the right number of robots in place to optimize the warehouse processes. Too many robots means that they are under-utilized and should be redirected to a place where they contribute more. Too few robots means that the warehouse is not going to fill as many orders as it could. So they have to get this just right.

But they don’t want to rely on the Goldilocks approach of trial-and-error to achieve that result. So they work off a great deal of data to track usage needs and make accurate predictions for how many robots a warehouse will need. The contract for services is based on that expectation, though a business is not forced to stick with that.

If needs change, robots can be removed or added (as they frequently are near holiday rush time now) to meet the customer’s requirements. The robots can be delivered within days, as they are shipped from California — where they are built — and fit into regular containers.

Voorhies envisions automating that process, as well: “Our dream is the robots load themselves into the shipping crate,” to be delivered where needed just as they deliver the boxes of products within the warehouse.