Robotics applications such as construction require specific robots such as SAM. Source: Construction Robotics
Thanks to artificial intelligence and machine learning, robots have become more aware and more autonomous than ever. They’ve already transformed many industries, from manufacturing to retail. Before robots can enter new markets, however, developers need to hone their software and hardware for the specific robotics application. A generic camera, gripper, or program might not be suitable, and a robot might even need to be redesigned to complete certain tasks.
Robotics engineers and suppliers have have done some amazing things, but we still have a long way to go before users can expect ubiquitous intelligent systems. This begs the question: Which robotics applications can benefit the most from new technologies? How are AI-enabled systems improving working conditions or providing other benefits?
1. In the operating room
Surgeons have always needed steady hands, as even the slightest of movements could cause them to nick arteries, sever tissue, or even pierce organs. There are no fully autonomous robot surgeons, but several systems have been developed to augment human capabilities.
Some examples of surgical robotics applications include Preceyes‘ system for ocular procedures; Corindus Vascular Systems’ CorPath system, which incorporates virtual reality (VR) for remote operations, and the Monarch robot from Auris Health.
A recent trend has been the integration of new technologies with advanced software for greater precision. With Preceyes, for instance, developers need to craft not just the surgical robot, but also the tools for remote operation with minimal lag or distortion.
The Preceyes system is designed for delicate eye surgery. Source: maxon precision motor
2. Law enforcement and emergency response
First responders, incident-response teams, and law enforcement officers increasingly rely on robots to go ahead of them in hazardous situations. The 2011 nuclear disaster in Fukushima, Japan, prompted advances in disaster-response robotics that continue to the present day.
Bomb-disposal robots, for example, are designed to assess threats and include a variety of sensors and remote monitoring tools. Similar robots can be used in emergencies navigating a burning building or a compromised structure after an earthquake.
The idea of a mobile platform is nothing new, but it is advancing with improved sensors, user interfaces, and autonomy. Human workers can do much more than simple monitoring. Ordnance-disposal robots such as QinetiQ’s Talon include manipulator arms and tools to precisely sever wires, for example.
Unlike remotely operated surgical robots, developers need to design field robots for unpredictable environments and to deal with large objects that could threaten their own safety.
How will a search-and-rescue robot or its operator know whether moving a concrete pillar that has fallen on someone will help or hurt that person? How much visual, location, or audio data does it need to gather for good decision-making?
Developers of such robots should work closely with professionals with experience out in the field or have similar experience themselves. As with any mission-critical technology for life-and-death situations, be prepared to conduct extensive prototyping and testing.
3. Welding with robots
Like surgeons, welders must be incredibly accurate with their movements. Keeping steady without distractions is tough enough, but throw in heavy welding gear — including the helmet — and it becomes a challenge of strength and precision. Partly as a result of this, welding is one of several industries in store for a labor shortage in the coming years.
Luckily, welding is another area that can benefit from robotics, not just on the factory floor but also out in the field. Robotics applications have been used in high-production environments for years, and it’s not uncommon to find dozens — if not hundreds — of robotic welders in automotive factories. As robots become more useful in construction, developers will have to tailor welding systems for the field. Ease of use will remain a key factor.
A joint venture between Universal Robots and ARC Specialties has resulted in two collaborative robot models for welding, one of which is called SnapWeld. It’s a water-cooled torch that can weld up to 600 amps. It includes a torch bracket, cables, and hoses. Similar to 3D printers, the robot can be fed certain parameters to perform on-the-fly stitch welding.
Welding robots for building or repairing bridges or other structures will have to be more portable and aware of their surroundings than stationary machinery. There’s also the issue of maintaining secure control in more open environments.
4. On the farm
Agriculture has always been subject to the vagaries of weather, soil conditions, pests, and the labor supply. Farming equipment has become much more sophisticated and can autonomously navigate, plant, weed, and harvest certain crops. Milking machines in dairies are just the beginning of robotics applications for livestock.
Additional robots handle duties like weed control, plant nursing and feeding, pollution monitoring, and even planting or seeding new crops. Demand has been high for robots to pick and move fruit, partly because of labor shortages.
Robots, drones, and deep learning are helping farmers to ensure their yields via precision agriculture. For example, Small Robot Company’s Tom uses a combination of GPS, AI, and mobile technology to move safely and digitally map fields. Despite Tom’s all-terrain wheels, its lightweight 3D-printed design means it won’t compact soil like most tractors.
Developers of robots and drones for precision agriculture need to understand analytics and big data systems. The robots will need good data collection and processing at the edge, good connectivity for sharing data, and even be able to tap into one another and the cloud for analysis and instructions.
Even then, human supervision is still necessary. Even a slight miscalculation could result in acres of crops destroyed in market where profit margins are slim, food safety is regulated, and robotics applications are still proving their value.
5. On the construction site
Robots applications in the construction and property development industries are being significantly enhanced thanks to technologies like AI, big data, and 3D printing. Australia-based Fastbrick Robotics can even build brick houses four times faster than human workers. It combines a 3D printer and a robot that can lay bricks just as precisely as human laborers.
Victor, N.Y.-based Construction Robotics‘ Semi-Automated Mason, or SAM100, is a collaborative bricklaying robot, and its Material Unit Lift Enhancer, or MULE135, is an assistive device for lifting up to 135 lb.
Other robotics applications include demolition, modular and on-demand construction, transport, surveying, and infrastructure inspection. Again, developers are integrating robots with other devices, such as drones, to further increase their usability. Imagine drones that can be used to transport materials or, better yet, ones that can build solid architectural structures.
There are still many challenges with developing rugged, reliable, and precise hardware and software for construction, which is one reason why there aren’t yet many robots in this space. 3D printing can be lucrative, for example, but it had to handle new materials such as concrete and scale up to create usable structures.
6. On the battlefield
Military robotics is unusual in that all of the application described above also could help military personnel. In addition to ordnance-disposal systems and drones for forward observation, armed forces around the world are interested in portable and easy-to-use surgical systems and rapid construction.
At the same time, aside from the obvious futuristic robot soldiers, some robotics applications are exclusive to the military. Last year, the British Army tested a variety of mobile weapons platforms, unmanned systems, and assistive technologies. Both the U.K. and the U.S., among others, are looking at improving autonomy, collaboration, and coordination among multiple systems.
MAARs is a modular, mini tank of sorts that can be outfitted with a variety of weapons and controlled remotely. Similarly, DOGO is a less intimidating robot that houses a weapon — such as a Glock — and camera for remote reconnaissance.
Note that many of these military robots are very hands-off. They require sensors that can deliver rich data in real time, clear user interfaces, active and secure connections, and a clear chain of command for deployment of weapons alongside human forces.
One advantage for developers is that the U.S. Department of Defense and its international allies and rivals are all willing to invest in game-changing technologies.
Designing to specific robotics applications
Although each of the robotics applications above is unique, there are certain shared goals and component technologies for developers to keep in mind. For example, in surgery, welding, and construction, robots need to relieve humans of the cognitive or physical burdens while improving precision and reliability.
In disaster-response, law enforcement, and military applications, robots may have greater autonomy because they are intended to go ahead of or instead of human professionals. Different sensors, power sources, end effectors, and methods of propulsion will be needed for different environments and use cases.
From farms to hospitals and building sites, robots and AI are changing how people work, so designers and testers should be keenly aware of the intended applications for each robot rather than lose focus by trying for a “one size fits all” solution.