Technology Trends That Will Propel Industries into the Future
With technology moving at a rapid pace, certain innovations are bound to change the way entire industries operate. Artificial Intelligence is gaining ground and becoming more and more advanced, and voice-enabled devices have been embraced over the last 2 years. Having seen the way things advanced, at this point in time we are able to see the bigger picture and figure out what changes are bound to come — some of them sooner rather than later. Each advancement will provide a bit of insight as far as how our lives will change.
It’s been a while since new smartphones that get released make any huge impact. Some studies show that people hang on to their older smartphones for longer period of times, due to the fact that new releases don’t bring to the table anything truly revolutionary. This can change with the foldable devices that are hitting the market. During the Samsung Unpacked event that took place this year in February, the new Galaxy Fold was announced. In addition, Huawei also unveiled their Huawei Mate X at the Mobile World Congress.
The Galaxy Fold combines the idea of having a small tablet as well as a smartphone into a single device. The device opens into a 7.3-inch tablet, and Samsung made sure to highlight the innovations made around software usability to allow for a seamless switch between phone and tablet modes. The company is a big player when it comes to the smartphone realm, which means that this trend could catch on, with more vendors possibly putting out their own versions of a foldable phone.
However, right now the new foldable smartphones are no mass market products, but luxury products. The Galaxy Fold wears a price tag of $2,000, while the Huawei Mate X has a price of over $2,500, which is a new record for smartphones. At the moment, having a foldable phone is seen as more of a status symbol, and only a small group will be in the market for this first generation of foldable devices, due to their exorbitant price tags, but it’s important to note the fact that it’s the first time in almost a decade that such a radical innovation took place in the world of smartphones.
Artificial Intelligence is being used all over the globe nowadays, with companies using deep learning and machine learning in order to improve the way data is analyzed and how problems are solved. At the moment, however, the process of setting up the infrastructure necessary for Artificial Intelligence and managing the lifecycle of a model is not at its easiest point.
In order for things to improve, cloud companies are currently creating new machine learning services, and new startups are popping up in order to offer necessary services in the field of AI — such as Paperspace, which provides tools required for training neural networks, exploring data, as well as running GPU compute jobs. More companies are expected to pop up in the AI category, offering tools designed to improve both model creation as well as management, the same way a set of processes and tools were designed for software development lifecycle in the software industry.
Other vendors that are currently offering their services are CognitiveScale and H2O.ai. CognitiveScale provides an Artificial Intelligence development platform which manages the lifecycle of Artificial Intelligence application, from development to production. Meanwhile, H2O.ai provides an open-source machine learning platform that uses linear scalability.
Robots are being used all over the world, but they are generally working in environments where humans aren’t present. The California-based tech company Nvidia announced in January of this year that they opened a robotics lab in Seattle in order to university researchers and talented people within the tech industry. The goal of the newly opened lab is the development of robots that are able to work alongside people.
Dieter Fox, professor at UW Paul G. Allen School of Computer Science and Engineering as well as the senior director of robotics research at Nvidia, noted the fact that in the past, research around robotics focused on small projects, as opposed to systems that are fully integrated. The lab is bringing together a team of experts in computer vision, deep learning, robot control and perception, and human-robot interaction. A team of 50 people are performing research in those areas, and in order for the research to remain relevant to the real-world robotics problems, the lab aims to investigate its work in the context of realistic scenarios when it comes to interactive manipulation. Dieter Fox noted that they feel that this is the right time to develop the next generation of robots.
One of these real-word scenarios that the research team focused on is a real-life kitchen, with a mobile manipulator solving a number of tasks — such as learning how to clean the table, retrieve objects and help cook a meal. The first manipulation system was demonstrated at an Open House event which took place in January.
The mobile manipulator shown at the event integrates state-of-the-art techniques in order to keep track of the drawers and doors in the kitchen, detect and keep track of the objects within the space, and gain access to certain objects. These techniques can be applied in random environments, but they require 3D models of cabinets and objects.
The robot uses deep learning, building on the company’s expertise when it comes to photorealistic, physics-based simulation, being able to detect specific objects based on its simulation, which means that no manual data labeling is necessary. The highly parallelized GPU that Nvidia uses allows the robot to keep track of the environment in real-time and adapt to changes that occur, by using sensor feedback.
The robot is performing real-time inference when it comes to processing and manipulation on the Titan GPUs, and uses the Jetson platform when it comes to navigation. cuDNN-accelerated PyTorch was used as the deep learning framework in order to train the robot ‘s perception system. The robot integrates a number of technologies that were developed by the researchers, which work together to enable tracking, detection, and to generate control commands. The robot’s system is built on and comprised by Dense Articulated Real-Time Tracking, Pose-CNN 6D Object Pose Estimation, Riemannian Motion Policies for Reactive Manipulator Control, and Physics-based Photorealistic Simulation.
IBM unveiled at the recent consumer electronics show that took place in Las Vegas, the IBM Q System One, which is a quantum computing system built for scientific as well as commercial use. In addition, the company also announced partnership as well as proof of Concepts with a number of large players from various industries. IBM was joined on stage by Exxon Mobil and discussed the way quantum computing could address problems that are computationally challenging across various applications. Using the power of quantum computing, Exxon could potentially optimize the power grid of an entire country.
The financial giant, JP Morgan Chase is currently working the IBM Q Network, focusing on cases where quantum computing can provide help within the financial industry — from risk analysis to asset pricing to trading strategies and portfolio optimization. The German multinational automotive corporation Daimler AG is also working with IBM in order to advance the potential of quantum computing to be used in the automotive and in the transportation industry.
When it comes to quantum computing, IBM is not the only company that focuses on it. Google, Microsoft, and even startups such as Rigetti are currently researching and also designing various aspects that pertain to quantum computing. Intel also has a hand in quantum computing, with Intel Labs showing the progress they made in the field at the American Physical Society March Meeting that took place in Boston. They presented not only the progress made towards the development of a quantum computing system that is commercially viable, but also pointed to breakthroughs made when it comes to hardware as well as software development for quantum computing stack. The development of a tool designed to speed up the research and the optimization of technologies referring to silicon spin qubit has also been highlighted.
At CES, Intel showcased Foveros, which is a 3D packaging technology, bringing together the benefits that 3D stacking has to offer, in order to enable logic-on-logic integration. This technology provides high-density, high-performance, and low-power silicon in a small package, which allows device manufacturers freedom to build a variety of products. Foveros is expected to be launched later in 2019 — in the second half to be more precise. The company announced that the Foveros products are ready for mass-market production, not only customized or specialized processors, but mainstream CPUs.
Given the pace that the tech field moves at, more trends will emerge in the years to come, and these are only a few areas that tech vendors are currently working on in order to come up with new improvements. Other areas that are also getting attention are robotic process automation, intelligence and mixed reality, and revenue optimization.
