What is behind trends like Digital Twin or Autonomous Shipping? The 17th Conference on Computer and IT Applications in the Maritime Industries (COMPIT) will give an insight. Organizer of Tutech Innovation presents an exclusive preview of the conference held 14–16 May 2018 in Pavone, Italy
Over the past two decades, COMPIT has established itself as a key conference in information technology (IT) for the maritime[ds_preview] industries, bringing together software developers and users. Most participants come from the industry, reflecting the practical relevance of the event. COMPIT lasts three full days. This year, the main trends, in short, are: • Appgrades galore: Do the same tasks, but better thanks to new technologies, using apps instead of installed software, smartphones instead of 3D scanning, or Virtual Reality instead of 2D displays for training and instruction. A lot of the progress is evolutionary. • Cradle-to-grave model-based processes: From early design to ships in service, product life-cycle management is model-based. The Digital Twin grows and evolves as the real ship does. • AI embraced: Assorted new developments use the power of Big Data and Artificial Intelligence (machine learning). AI is embraced as a useful tool, nothing more, nothing less. • Autonomous systems spread: While everybody waits for the unmanned ship on the horizon, under our noses a multitude of autonomous systems and robots mature to support smarter ship operation. Tear your eyes away from the horizon and grab the opportunities! Appify me If computer experts would speak plain English, we all might understand them. Which is probably a nightmare for them and will not happen in this world. So, us mere mortals better get used to some new vocabulary. »Appification« for example. The act of turning software that used to be installed on your computer into an App – software that is in the cloud and can be run from your smartphone. Actually, that is quite a smart idea and the Apps are coming also to the marine world. »Appify what can be appified« seems to be the motto this year. Palluch et al. (Friendship Systems and ISA Propulsion) describe a web-based App (webApp) for geometric modeling and design of propellers. The App builds on well-established design environment CAESES offering selected functionalities via a standard web-browser. Building on CAESES’ parametric modeling techniques, a propeller is generated with just a handful of inputs. Its geometry can be downloaded in standard format descriptions, e.g. for CFD simulations or 3D printing. There are persistent rumors that smartphones are actually sometimes used for phoning other people. But surely, they are busy most of the time with these Apps or other engineering wizardry. Did you know that smartphones can replace 3D scans? This little gem is presented by Fischer et al. (Fraunhofer and Rostock University), who describe new ways of capturing the as-built state of ships. Instead of expensive laser scanners, mobile devices like smartphones are used to create point clouds and ultimately 3D models. During assembly, the 3D model of the ship as-built is then available with precious little time delay. Such 3D models are the backbone of a host of applications from early design to maintenance of the life-time of a ship. One of the most popular applications uses 3D visualizations in Virtual Reality (VR) or Augmented Reality (AR). Increasingly realistic computer generated images can be used for making training and instructions more intuitive and user-friendly, ultimately saving time and cost. VR training applications are akin to video games – you need to create a sufficiently large virtual world to make it interesting, with sufficient level of detail and behavioural realism and with sufficient interactivity and training elements for the player/learner to keep him playing/learning. And that comes at a hefty price. Did you know that high-end video games cost double-digit millions to produce? Even if your expectations are not quite that ambitious, production costs are still often prohibitively high for our industry. This explains why there are so few maritime VR training applications on offer, as found by Bertram & Plowman (DNV GL) in their survey. Friedewald & Meluzov (TU Hamburg-Harburg) might offer a silver lining on the horizon with their work on bringing development time and cost down, in their case for Augmented Reality based instruction modules as intended for ship services and maintenance. One Digital Twin to bind them all The »Digital Twin« at least sounds like proper English. And Google gives me four times as many hits as for Appification. So, a lot more people use this expression, but what is it? A 3D geometry model of a ship is a CAD model. Add information about steel type, suppliers, inspection dates, etc. and you get a PDM (product data model). Add models that mimic the behaviour (strength, vibration, hydrodynamics, etc.), and track the changes of this behaviour over the lifetime of the ship, and you get a Digital Twin. In essence, the vision is to have a computer model with the look and feel of the real deal. As with Virtual Reality, the vision is much grander than the reality of current implementations, but what we have now is already useful. Morais et al. (SSI) point out that the »Digital Twin journey« is a rocky one: »While the benefits of a digital twin are real, achieving the desired outcome is more challenging than generally thought. Many obstacles, some cultural, some process related and some technological, commonly appear when trying to create a digital twin.« Niepert & Grau (Prostep) show how a Digital Twin could be used in practice to accelerate design and production, describing model-based class approval and site inspection. Historically, these processes required numerous drawings, transferred between classification society and shipyard. Much time is saved if 3D models are enhanced by functionalities such as digital signatures, annotations that convert directly into data-management systems, search functions, or digital highlighting. With 3D PDF, you can create fully customizable 3D documents that serve this purpose. And like 2D pdf documents, the software to read these documents is widely available and free of charge. Aye, aye, A.I. One of the often used and rarely understood buzzwords of recent times is Artificial Intelligence (AI). The time is ripe for demystifying it. Bertram (DNV GL) gives an introduction to key AI technologies (machine learning, expert systems, speech and gesture recognition) illustrating each technique with maritime examples. The take-home message here is that there is no need to shiver in awe or fear that humankind will be eliminated by a robot revolution. AI is rather an evolution of engineering tools just as moving from polynomials to spline curves was when CAD was in its infant stage. »Just an engineering tool«, but a powerful one, and we would be stupid not to embrace the evolving power of Big Data and machine learning. Perez (Sener) outlines his vision for Big Data and AI based detailed design for ships, where full-scale performance data are piped back into design rules which supplement simulation-based first-principle design. As Perez admits, »some of these improvements may seem unrealistic in the short term, but reality often exceeds expectations«. And each journey starts with first steps towards the vision on the horizon. Woo et al. (Korea Maritime & Ocean University et al.) have collected production data from several Korean shipyards and harvested this wealth of data to improve the standard information system for the time factor of process lead-time. Xie et al. (DNV GL) use artificial neural nets to find automatically cracks in the ship structure on images from drone-based inspections. Most applications of Big Data and machine learning emerge for ships in operation, where AIS data is often used for other purposes than the original purpose of increasing safety in shipping. Cepeda et al. (COPPE) use AIS data to estimate ship emission in the port of Rio de Janeiro. Holm (RISE Viktoria) adds an interesting element for anybody working with AIS data. His approach speeds up decoding large AIS encoded data sets into clear text by a factor 20-30 compared to alternative decoding. Good news for performance monitoring and other applications! Autonomous under your nose There is no escape from the discussion about unmanned shipping – also not at COMPIT. However, while transatlantic 20.000TEU containerships without crews still are utopia, Artificial Intelligence enters in many forms our daily life, also in marine operations. Think small, think components – and you will see it. Situational awareness and collision avoidance is such a component where autonomous technology progresses rapidly from concept to implementation. The developments are in many ways akin to close-range parking aids in cars: While the driverless car makes the news, systems for enhanced situational awareness and support for parking are so ubiquitous in modern cars that we hardly notice them at all anymore. Le Gallic et al. (Aland University) test autonomous collision avoidance on a 4m autonomous surface vessel. Using such ASVs for prototype testing of algorithms is typical, as test platforms are cheap and legally much easier to operate than real ships. Alternatively, you may test in nautical simulators, such as Schubert et al. (University of Applied Sciences Wismar) or Tank et al. (University of Oldenburg). While many smaller autonomous systems become already reality, some nations set sight on the horizon where we see unmanned ships arise. Norway is one of them, but Finland is in hot pursuit. One Sea, a consortium of industry partners, is paving the path to autonomous maritime operations, as described by Merenluoto (Dimecc). Matsuo (NMRI) seems to be determined that Japan would be at least among the medal ranks. The Japanese roadmap to shipping 2050 foresees unmanned shipping side-by-side with manned ships – where crew training is replaced by direct knowledge download from the cloud to the brain. Until we have such direct downloads as a convenient highway to insight and knowledge – a Nuremberg funnel of the digital ages – the next-best thing is attending COMPIT with an open mind for new technologies and contacts.
