AI is certainly on the rise, despite some of the concerns some have expressed about it leading to doomsday scenarios or a devastating loss of jobs, etc. When it comes to AI and enterprises, many have already begun to implement AI as a part of their business and digital transformation strategy. In fact, about 80% of organizations are already using AI in some form. However, an overwhelming 91% of companies foresee significant barriers to AI adoption, such as a lack of IT infrastructure and a shortage of AI experts to guide the transition. Nearly half of the tasks currently undertaken by humans could already be automated, even at current levels of technology.

In this paper, we introduce an actor-critic algorithm called Deep Value Model Predictive Control (DMPC), which combines model-based trajectory optimization with value function estimation. The DMPC actor is a Model Predictive Control (MPC) optimizer with an objective function defined in terms of a value function estimated by the critic. We show that our MPC actor is an importance sampler, which minimizes an upper bound of the cross-entropy to the state distribution of the optimal sampling policy. In our experiments with a Ballbot system, we show that our algorithm can work with sparse and binary reward signals to efficiently solve obstacle avoidance and target reaching tasks. Compared to previous work, we show that including the value function in the running cost of the trajectory optimizer speeds up the convergence. We also discuss the necessary strategies to robustify the algorithm in practice.

--The increasing penetration level of energy generation from renewable sources is demanding for more accurate and reliable forecasting tools to support classic power grid operations (e.g., unit commitment, electricity market clearing or maintenance planning). For this purpose, many physical models have been employed, and more recently many statistical or machine learning algorithms, and data-driven methods in general, are becoming subject of intense research. While generally the power research community focuses on power forecasting at the level of single plants, in a short future horizon of time, in this time we are interested in aggregated macro-area power generation (i.e., in a territory of size greater than 100000 km Real data are used to validate the proposed forecasting methodology on a test set of several months. A. Motivations As the penetration level of Renewable Energy (RE) sources is growing worldwide to meet ever tightening sustainability goals [1], the intermittent and uncertain nature of RE is posing increasing challenges to efficiently manage a power grid, eventually endangering its own stability. In this context, the availability of accurate forecasts of power generation from RE may mitigate the impact of the increasing penetration level and improve the operation of power systems [2].

Terra Drone Corporation has launched a new UAV and AI-based solution set for the maintenance of power transmission and distribution equipment. The solution was developed based on the market gaps identified after inspecting more than 90,000 km of power lines by BVLOS throughout the world. Acquired data is automatically processed and analyzed by artificial intelligence algorithms, which are trained to detect crossovers at the bottom of transmission lines, buildings and construction machinery. The system identifies rust on bolts, loosening and missing tower parts, bird's nests and vegetation encroachment and generates a smart report, highlighting the areas that require action. The error (identified anomaly) detection system is accurate up to 92.5%.

The Energy Department unveiled plans to launch a $5.5 million research center that will bring together top thinkers from three federal and academic institutions to solve some of the world's most complex challenges in artificial intelligence. Researchers from Energy's Pacific Northwest National Laboratory, Sandia National Laboratories, and the Georgia Institute of Technology will join forces in the newly unveiled Center for Artificial Intelligence-Focused Architectures and Algorithms, or ARIAA. "AI will allow us to solve problems today, that simply cannot be solved because they are too complex," said Roberto Gioiosa, the senior research scientist at PNNL selected lead the new center. "This is the science of the future." The move follows Energy Secretary Rick Perry's recent announcement that the agency is working to establish an Artificial Intelligence and Technology Office to coordinate and streamline the agency's efforts and growing investments around AI.

Whether businesses are using machine learning to perform predictive maintenance or create better retail shopping experiences, ML has the power to unlock value across a myriad of use cases. We're constantly inspired by all the ways our customers use Google Cloud AI for image and video understanding--everything from eBay's use of image search to improve their shopping experience, to AES leveraging AutoML Vision to accelerate a greener energy future and help make their employees safer. Today, we're introducing a number of enhancements to our Vision AI portfolio to help even more customers take advantage of AI. Performing machine learning on edge devices like connected sensors and cameras can help businesses do everything from detect anomalies faster to efficiently predict maintenance. But optimizing machine learning models to run on the edge can be challenging because these devices often grapple with latency and unreliable connectivity.

There different approaches and attempts to decentralize or distribute various aspects of computing. In most cases, it is tried to distribute some customer use cases, whereas the underlying algorithms are so well known that an individual result validation can be developed. However, if a distributed computing network only focuses on supported use cases or requires customers to also supply a validation method, the usability is heavily limited for customers. HiveNet on the contrary focuses on mass adoption with flexible use cases. Customers will be able to request the computation of a wide set of tasks, which they can define, without the need of additionally supplying validation methods.

Roundup Welcome to this week's machine learning musings. Google has upset city officials by trying to improve its facial recognition technology, and the new TensorFlow 2.0 has been released. Google offered $5 gift vouchers to black homeless people and Atlanta city isn't happy: Facial recognition datasets are unfairly dominated with images of white men, so Google hired third-party contractors to go around recording people's faces by offering them vouchers. The temp agency, Randstad, were told to target people of darker skin, and, unfortunately, some of those people were homeless people living on the streets in Atlanta. The methods used to tempt them were ethically dubious.

Generative Adversarial Networks have been shown to be powerful in generating content. To this end, they have been studied intensively in the last few years. Nonetheless, training these networks requires solving a saddle point problem that is difficult to solve and slowly converging. Motivated from techniques in the registration of point clouds and by the fluid flow formulation of mass transport, we investigate a new formulation that is based on strict minimization, without the need for the maximization. The formulation views the problem as a matching problem rather than an adversarial one and thus allows us to quickly converge and obtain meaningful metrics in the optimization path.