Listen to this article
Over the past few years, the internet has closely followed along as Boston Dynamics’ Spot has more or less “grown up.” We’ve collectively watched in awe as Spot and its robotic predecessors learned to withstand a push or a shove, navigate dangerous terrains, and eventually dance like BTS.
For the general public, Spot’s robotic dance montages are what the robot is best known for. But is Spot more than a neat (albeit expensive) commercial toy? The commercial drone market provides one answer.
The Path to Drone Development
In 2016, the analyst firm Gartner predicted that commercial UAVs were just barely rounding the corner of “peak inflated expectations”, with 5-10 years of development still to come before the technology reached mainstream adoption. By 2017, Gartner concluded that drones had effectively rounded the corner, with the adoption time frame essentially halving to 2-5 years.
Quite rapidly (though less publicly than the perception of ground robots), drones went from being freelance videographer toys and neighborhood nuisances to essential inspection tools for enterprise businesses like construction, energy and agriculture. The number of use cases for drones also grew as drone hardware became more advanced and reliable. Advanced aerial mapping, crop monitoring, public safety uses, disaster response, and consumer drone deliveries all seemed to be on the table as the drone industry matured.
Today, there are over 350,000 registered commercial drones in the U.S., with some of the largest fleets, such as that of agricultural giant Corteva, numbering in the hundreds. Adoption has been primarily driven by the various safety and efficiency benefits that drones provide to industrial jobsites.
These hundreds of thousands of drones are now conducting programmed crop scans, completing exterior building inspections, and performing thermal scans of solar farms with routine swiftness. They are keeping workers safely away from disaster areas or dangerous construction tasks (i.e., inspecting roofs, scaling scaffolding), saving time, resources and human well-being. However, the industry would not have gotten to this point without the help of advanced software solutions.
The merger of hardware and software is broadening and extending ground rovers’ abilities. Without it, ground robots tend to be stuck at the stage of pure navigation rather than moving toward data capture, analysis, and eventually, actuation.
Robotic Software Has Its Moment
This month, Google parent company Alphabet announced its newest foray into the field of robotics with Intrinsic. Intrinsic, set to “unlock the creative and economic potential of industrial robotics” through software tools, is a substantial departure from some of the company’s more hardware-focused efforts, such as the acquisition and eventual sale of Boston Dynamics. This news indicates Alphabet, similar to many companies over the years that have more or less abandoned their robotic hardware ambitions, is now seemingly placing a bet on software.
And just in time. For aerial robots, advanced software solutions that leverage AI and machine learning are not new – digital twin and mapping technologies, have been on the market since the early 2010s and have since only grown. But it is only recently that this software has begun to power ground rovers and interior robotics in industrial settings.
Last year, leading construction firm Brasfield & Gorrie began outfitting Spot with cameras and DroneDeploy, software for generating aerial maps and models, to capture ground-level imagery and gain unlimited visibility into the state of their jobsites. The merger of such hardware and software is broadening and extending ground rovers’ abilities. Without it, ground robots tend to be stuck at the stage of pure navigation rather than moving toward data capture, analysis, and eventually, actuation.
When businesses talk about the future of robotic automation, they tend to talk about actuation. Actuation, in an industrial setting, is automated tools and software platforms working in conjunction to capture, analyze, and automatically act upon gathered data in a “closed-loop” system that requires minimal human intervention or oversight.
The Future is Actuation
When businesses talk about the future of robotic automation, they tend to talk about actuation. Actuation, in an industrial setting, is automated tools and software platforms working in conjunction to capture, analyze, and automatically act upon data gathered in a “closed-loop” system that requires minimal human intervention or oversight. As one example, consider a pre-programmed drone scanning a farm every morning, detecting problem areas, and automatically activating a ground robot to spray/address the issue. Essentially, the software platform powering the robotic data capture and analysis would also power the resulting action.
This vision is perhaps still a few years away. Still, several companies are already progressing toward this goal as developments in robotic hardware reach an inflection point. Alphabet’s Intrinsic has admitted that the “surprisingly manual and bespoke process of teaching robots how to do things…is currently a cap on their potential.” It is for this reason that at DroneDeploy we have taken a step toward a future of actuation with the acquisition of Rocos, a New Zealand-based provider of robot management software that allows companies to build, test, deploy, automate, and coordinate the operations of robot fleets.
When interior and exterior data capture is integrated on a single analytical platform, complete workflow automation and actuation for aerial and ground robots becomes achievable. While it may be nice to watch Spot dance for now, it will be even more impressive to watch Spot – perhaps in concert with a drone – act on its own to repair a solar panel or repair a crack in a building in just a few years.
About the Author
Nicholas Pilkington is the co-founder and CTO of DroneDeploy, the leading enterprise drone data company. Nicholas leads technical strategy at DroneDeploy and is responsible for innovation and research leadership. Prior to DroneDeploy, he completed his PhD in Machine Learning at the University of Cambridge. He also holds an M.Phil in NLP and Signal Processing and BSc in Computer Science, Applied Math and Information Systems.
- RoboDK – Start-Up Profile – Powerful, Cross Platform, Robotics Simulation & Programming Environment
- Start-up Profile – Southie Autonomy – Simplified Industrial Robotics Programming Via a Gesture-based Interface
- Start-up Profile – Omnirobotic – Reducing the Cost of Robot Programming for ‘High Mix’ Manufacturers
- Prophesee Debuts Metavision Intelligence Suite SDK as Part of Ecosystem for Event-Based Visioning
- TriVision A/S – Start-Up Profile – Machine Vision Solutions for Food and Packaging Inspection
- Start-up Profile – DreamVu – Novel 360° 3D Vision System Enables VR, Localization, Mapping, Object Tracking, More