'The Deep Learning Summit brought together a fantastic collection of people from diverse areas of the field, from both academia and industry, with a balance of both technical and industry focused backgrounds’' Drew Volpe, First Star Ventures (Deep Learning Summit, San Francisco, 2017)
SAN FRANCISCO (PRWEB) April 24, 2018
These events will bring together 550 experts and 60 speakers working in AI and Deep Learning. All attendees are welcome to attend presentations and workshops from both tracks, providing an in-depth but varied agenda. At the Deep Learning for Robotics Summit, global experts will explore where AI meets the real world through human-robot interaction, imitation learning, skill embeddings, robot perception and many other areas. The AI in Industrial Automation Summit will delve into industry 4.0 and discover how to harness advancements in AI to create factories of the future.
Manufacturing is susceptible to bottlenecks in production, inefficiency, waste and high costs in production. The Industrial Internet of Things (IIOT) is bringing us the factories of the future with the implementation of AI to help optimise and scale operations through predictive maintenance, improved safety amongst others. Digital technologies promise to transform the industry and there’s a huge buss around advanced analytics, augmented reality, additive manufacturing, and other technologies. However, many companies struggle to create value. Greg Kinsey, Vice President of Industrial Solutions and Innovations at Hitachi will share his expertise in creating practical value from AI in manufacturing. He explains that "some companies fail by focusing purely on technology", and will explain how AI can be used to predict and avoid defects and bottlenecks, and well as presenting a roadmap for digital transformation of manufacturing.
Another company leading the way in this area is Caterpillar, and Analytics Tech Specialist, Benjamin Hodel, will share breakthrough research on how they’re creating better earth-moving machines controlled through AI. Developing products is expensive, and virtual product development is able to iterate new designs in simulation before prototypes are built, avoiding unnecessary expenditure. Benjamin explains that when the simulations are run by a computer, the machines need to be able to operate in a "human-like" way so that the results can be trusted. Caterpillar Inc. have identified challenges in creating these models, as the simulation environment is very dynamic, meaning simple rules fail to achieve the right behaviour. Through using reinforcement learning a program is able to learn to operate an earth moving machine by itself and learn to improve its behaviour over time.
Recent results in the fields of deep and reinforcement learning have also ignited immense interest in the possible applications for robotics, however there are several areas where robots are still falling short. Derik Pridmore, CEO of Osaro explains that "humans are able to learn new skills from interactions and experiences in the real-world. Whilst AI has already beaten the best human player at Go, and achieved superhuman performance in many video games, current AI systems using advanced DL techniques still can’t accurately navigate a car in the real world, even with millions of miles of driving data. A human, in this instance, would not need significant experience - humans learn new skills using a commonsense understanding of physical objects and intentional agents." Osaro is a machine learning company building products powered by deep reinforcement learning, and at the Deep Learning for Robotics Summit, Derik will discuss the most recent advances in this space and the current state of industrial robotics.
Both summits bring together industry and academia to share their knowledge and expertise, and presenting her work in Semantic Understanding for Robot Perception is Jana Kosecka, Professor and visiting research scientist at the George Mason University and Google. She explains how "advancements in robotic navigation and fetch and delivery tasks rest to a large extent on robust, efficient and scalable semantic understanding of the surrounding environment. Deep learning fueled rapid progress in computer vision in object category recognition, localization and semantic segmentation, exploiting large amounts of labelled data and using mostly static images.
"There are both challenges and opportunities in tackling these problems in indoors and outdoors environments relevant to robotics applications. These include methods for semantic segmentation and 3D structure recovery using deep convolutional neural networks (CNNs), localization and mapping of large scale environments, training object instance detectors using synthetically generated training data and 3D object pose recovery."
Additional confirmed speakers across both tracks include Andrei Polzounov, Senior Research Scientist at Blue River Technology sharing his work on Image Segmentation for Precision Agriculture; Alicia Kavelaars, Co-Founder and CTO at OffWorld who will be discussing an Industrial AI Revolution in Space; Yibiao Zhao, Co-Founder and CEO of iSee talking about Engineering Common Sense; Animesh Gard, Postdoc Researcher in Robotics and DL at Stanford University who will be discussing his research in Generalisable Limitation in Robotics, and many more.
Early bird passes end May 4th, so register now to guarantee your place at a discounted price. Join RE•WORK and and meet global experts in the field who will share their work in 15 minute presentations followed by a 5 minute Q&A, get involved with the interactive workshop track, and explore the exhibition space. This event will sell out.