The Galaxy S8 and Galaxy S8 Plus have been on the market for just over one month and continue to be a hot topic among smartphone enthusiasts. Whether it's Samsung celebrating favorable sales of the handsets or hackers attempting to spoof the device's iris scanner, there is still a lot to say about the Galaxy S8. The smartphone introduces to Samsung's line a new design and new software features, such as the AI assistant Bixby. It also touts several feature improvements, including a sharper and faster camera and more efficient processors. Whether the Galaxy S8 is worth the buy is completely up to the consumer; however, there are many things to take into consideration when selecting a new premium smartphone. The Samsung Galaxy S8 is a beautiful smartphone. It stands out during a time when mobile devices have become just ubiquitous enough for many consumers to not care about design, and would rather just have a functional device.

Fixed-wing drones are the way to go for efficient flying, but they pose challenges for long term autonomy because of how demanding they are when it comes to takeoffs and landings. You need a nice big flat area, and usually you need infrastructure support. A drone that needs to operate for days or weeks at a time completely on its own can't rely on either, which means you need to get creative. At ICRA this week, researchers from the University of Sherbrooke in Canada have gotten creative, and came up with a very clever design for a fixed wing drone called SUWAVE (Sherbrooke University Water-Air VEhicle) that uses lakes as landing pads. It crash lands in them, recharges with solar power, and then takes off again with a brilliant hinged propeller.

Summary: If you want to capitalize on all the amazing advancements in data science take a look at these two hot growth areas for IoT. It's likely that these will be where a lot of venture capital is invested over the next year or two. A lot of well deserved attention is being directed at speech, image, and text processing. The tools in this area are the CNNs and RNNs we've reviewed in recent articles. We'll continue to exploit and refine these capabilities probably for several more years but if you want to get out in front you really need to be looking for the next wave.

Logitech has announced a pair of new computer mice with the MX Master 2S and the MX Anywhere 2S. Both are updated versions of Logitech's previous MX models, and both come with a new "Flow" feature for multiple computer support. Logitech's MS Master and MX Anywhere were both considered to be the company's two best mice, and it just seems appropriate that both have new, upgraded models. Like with its previous iteration, the MX Master 2S's thumb wheel can be customized with additional functions through the Logitech Options software. As for the MX Anywhere 2S, it comes with the same speed adaptive scroll wheel feature.

DRONES are sending back snaps from the deep. Uncrewed vehicles that scour the ocean floor for submerged mines can now beam back images to human operators in close to real time. The technology, developed by Canada's Department of National Defence, could also be used to autonomously map the locations of starfish colonies, for example, or study deep-sea hydrothermal vents. The torpedo-shaped drones constantly scan the ocean floor using sonar, reaching speeds of around 2 metres per second. They use image recognition to search for shapes that look like a submerged mine.

A new startup launched today that seeks to streamline electricity purchases with the use of a peer-to-peer network that relies on artificial intelligence and other technology like high-frequency trading to balance supply and demand in real time. The idea behind the company, named Drift, is to remove inefficiencies from the current wholesale power market structure in deregulated service areas and pass the savings on to consumers. Drift is initially rolling out its service in New York City and plans to expand to other parts of the country this year. The company will operate like a mini Independent System Operator (ISO) for a distributed energy platform. Drift has nearly 3,000 independent power producers in its network that range from hydroelectric dams, to solar-plus-storage projects, wind farms and large commercial buildings.

Of course, the limitations of this is that it is a best efforts prediction by expert practitioners. As the case of solar energy predictions shows, insiders are among the most conservative in their own field. Google DeepMind has bested human players this year. It did so using a tensor processing unit, which uses one tenth of the power as the GPU set up last year. So undeniably progress is fast. AlphaGo also paired up with human Go players in an interesting example of intelligence augmentation.

What if an insurance company had a tool that improved profitability by predicting insurance rates and premiums with greater precision and efficiency? Or if a natural gas utility company improved safety, efficiency, and customer satisfaction with a tool to predict and better respond to leaks or unaccounted-for gas? These are both examples of machine learning techniques that are gaining momentum due to the pervasiveness of data (from the Internet of Things [IoT], social media and mobile devices) and the seemingly infinite scalability of cloud-based compute power. At the same time, the understanding and sophistication of algorithms have expanded, and the ability to apply complex mathematical calculations to data, and rapidly process them, is also driving growing interest in exploiting machine learning to gain competitive advantages in business. Machine learning is a type of data analysis technology that extracts knowledge without being explicitly programmed to do so.

Accurately predicting the future capacity and remaining useful life of batteries is necessary to ensure reliable system operation and to minimise maintenance costs. The complex nature of battery degradation has meant that mechanistic modelling of capacity fade has thus far remained intractable; however, with the advent of cloud-connected devices, data from cells in various applications is becoming increasingly available, and the feasibility of data-driven methods for battery prognostics is increasing. Here we propose Gaussian process (GP) regression for forecasting battery state of health, and highlight various advantages of GPs over other data-driven and mechanistic approaches. GPs are a type of Bayesian non-parametric method, and hence can model complex systems whilst handling uncertainty in a principled manner. Prior information can be exploited by GPs in a variety of ways: explicit mean functions can be used if the functional form of the underlying degradation model is available, and multiple-output GPs can effectively exploit correlations between data from different cells. We demonstrate the predictive capability of GPs for short-term and long-term (remaining useful life) forecasting on a selection of capacity vs. cycle datasets from lithium-ion cells.

This week we covered the reality of self-driving, startup positioning, space chips, fusion-powered future, and superintelligent physics. What driving through 25% of the United States feels like. Our year and a half living on the west coast of the US in southern California has come to a close. In order to mark the occasion, we are driving coast-to-coast across the US from San Diego to Manhattan. We've met up with business partners, parents, and puppies along the way.