5 AI Technologies to Watch
Tech buzzwords are typically vague — “AI” and “IoT” especially so. Here, we take a look at five technologies that intersect both realms that have a path, clear or not, toward commercialization.
August 2, 2019
![Image shows a drone over a city. Image shows a drone over a city.](https://eu-images.contentstack.com/v3/assets/blt31d6b0704ba96e9d/blt52b8b3558016c8db/63abe772eebe594a54c207ee/Drone-20.png?width=700&auto=webp&quality=80&disable=upscale)
Getty Images
Smart robots are an emblematic IoT technology by virtue of their connected intelligence. They also call to mind science fiction thanks to their ability to learn from (and sometimes interact and collaborate with) humans and sense their surroundings. While adoption of truly smart robots is still at an early phase, they have an array of potential applications — from automating nursing tasks to working alongside humans in factories, warehouses, restaurants and retail environments. The startup Knightscope has even debuted autonomous security guard robots. While over-the-top coverage of smart robots warns of their capacity to automate away countless jobs, adoption of truly smart robots pales in comparison to traditionally programmable robotics.
Although they theoretically threaten to displace millions of jobs in the form of truck, taxi and delivery drivers, autonomous vehicles could, in the near term, have an uneven effect on professional drivers. In trucking, autonomous vehicles could be an enabling technology, helping the industry fend off an acute shortage of drivers while shifting more human driving work to local delivery. Self-driving trucks will likely first be deployed on freeway-based long-haul routes. Research from the American Center for Mobility acknowledges the likelihood that autonomous vehicles could result in a loss in some passenger-based or car-based driving jobs, but adds that limousine, bus and transit drivers face less risk of having their jobs automated away in the relatively near term. While the disruptive potential of autonomous vehicle technology at large is fairly certain, the technology was clearly overhyped until fairly recently, after an experimental self-driving Uber ran over and killed a pedestrian in Arizona in March. Following that news, regulatory and social pressures have stiffened, chilling the formerly unbridled enthusiasm in the technology, which had been progressing quicker than many transportation experts expected. While investment has cooled, competition remains fierce, pitting tech companies such as Alphabet’s Waymo against legacy automotive companies ranging from Ford to Toyota. Meanwhile, some companies, including Toyota, are beginning to shift their focus on solely developing driverless vehicles to creating vehicles designed to maximize safety.
Virtual assistants seem to be everywhere these days, with Amazon’s Alexa functionality showing up in perhaps the broadest number of places: hotels, cars, smartphones, TVs and, of course, smart speakers. While not all experts agree with the nomenclature of referring to virtual assistants as conversational AI, the capabilities of such systems are steadily increasing, which could lead to such systems popping up increasingly in the enterprise, health care and industrial realms. One factor boding well for the technology is its relative ease of access for developers, who can use relevant toolkits to incorporate virtual assistant technology into a range of products and settings. Virtual assistant technology also has broad support from a range of powerful vendors including Apple, Samsung and Microsoft as well as Amazon. Although it is not always feasible to use voice commands in, say, noisy or crowded environments, the technology is also noteworthy for its potential to serve as a sort of unified user interface for IoT applications. In a successful interaction with a virtual assistant, a user could theoretically launch multiple programs with a single utterance. The technology is, however, decidedly in an experimental phase, with firms like Amazon continuing to unveil a steady stream of whimsical features such as the ability to yodel or do impressions. While the capabilities of such systems are steadily improving, their functionality in the time being is mostly limited to simple commands.
Digital twins are essentially the intersection of computer modeling with the Internet of Things, used to digitally replicate smart connected devices. The central selling point for digital twins, however, is not just their ability to use IoT-based sensor data to replicate the physical assets and systems, but their support for machine learning, which can, in turn, be used to help optimize the operation of assets or products — frequently in the industrial realm. Descriptions of digital twins often may have considerable overlap with computer models with decades of history, but digital twins are unique in their ability to accommodate real-time operational data and thus continuously “learn” as a result. While tech and industrial vendors often couch “digital twin” as an emerging technology, the concept dates back to the early 2000s.
While drones are most commonly associated with military, imaging and hobbyist applications, their capabilities are expanding thanks to advances in analytics and artificial intelligence domains such as computer vision. It is now feasible for drones to help assess property damage in the wake of storms, monitor crop growth in agriculture and streamline inventory management, as well as monitor construction and industrial sites including quarries and utility infrastructure. Owing to regulatory and infrastructure challenges, the potential for first responders to use drones to ferry medical supplies to patients in emergencies is less certain. The prospect of widespread use of drones in transportation and logistics in the near future faces similar hurdles. Google-parent Alphabet and Amazon have worked on delivery technology that could one day transform package delivery. But deploying the technology is certainly not as straightforward as Amazon leader Jeff Bezos hinted in a “60 Minutes” interview in late 2013 that an autonomous drone-based package delivery would become a mainstream technology by 2018. While Amazon apparently rushed the gun in promising it could deliver packages to customers in 30 minutes or less, expanded use of drones for logistics and other applications is a safe bet, although regulations will likely constrain use cases to defined applications.
While drones are most commonly associated with military, imaging and hobbyist applications, their capabilities are expanding thanks to advances in analytics and artificial intelligence domains such as computer vision. It is now feasible for drones to help assess property damage in the wake of storms, monitor crop growth in agriculture and streamline inventory management, as well as monitor construction and industrial sites including quarries and utility infrastructure. Owing to regulatory and infrastructure challenges, the potential for first responders to use drones to ferry medical supplies to patients in emergencies is less certain. The prospect of widespread use of drones in transportation and logistics in the near future faces similar hurdles. Google-parent Alphabet and Amazon have worked on delivery technology that could one day transform package delivery. But deploying the technology is certainly not as straightforward as Amazon leader Jeff Bezos hinted in a “60 Minutes” interview in late 2013 that an autonomous drone-based package delivery would become a mainstream technology by 2018. While Amazon apparently rushed the gun in promising it could deliver packages to customers in 30 minutes or less, expanded use of drones for logistics and other applications is a safe bet, although regulations will likely constrain use cases to defined applications.
On the surface, the collision of AI and IoT may sound like buzzword bonanza, but the prospect of bringing intelligence to physical objects has long been a core promise of the Internet of Things. You would be forgiven for thinking that technologies spanning both AI and IoT would be so hype-laden that they may never live up to the projections. But the following five technologies are very real and are in various stages of adoption.
About the Author(s)
You May Also Like