AI is maybe the most powerful tool our generation has available. Andrew NG called it "the new electricity". But what does it take to build AI enabled products? What are the key elements to achieve production grade AI? How does it impact your development process? How can quality be achieved? These are the questions this talk tries to answer. You will get an idea why the industry is talking about nothing less than a paradigm shift when it comes to developing AI based products.

Target Audience: Everyone interested in the shift from classical software engineering to data driven AI applications
Prerequisites: Interested in AI, how it works and its impact on engineering departments
Level: Advanced

Extended Abstract:
AI is maybe the most powerful tool our generation has available. Andrew NG called it "the new electricity". Most likely you used an AI based product within the last 3 hours, maybe without even noticing it. But what does it take to build AI enabled products? What are the key elements to achieve production grade AI? How does it impact your development process? How can quality be achieved? These are the questions this talk tries to answer. In addition we will look into the different stages of AI development and the tools which can help to make this process more efficient. You will get an idea why the industry is talking about nothing less than a paradigm shift when it comes to developing AI based products.

Whether evolution or revolution, or yet old wine in new skins, for more than 10 years, DevOps is changing how we develop and deliver software. This session looks back on the roots of DevOps, its movement until today, and current as well as possible future directions. This interactive session aims to offer a set of fruitful starting points for reflection and discussions.

Target Audience: Anyone interested in developing and delivering software
Prerequisites: Knowledge in DevOps and agile software development
Level: Advanced

Was wäre, wenn ein Brain Computer Interface (BCI) Ihre Gedanken in Befehle zur Steuerung von Software oder Hardware übersetzen könnte? Das TNG Innovation Hacking Team hat sich auf Basis eines OpenBCI mit dem Thema auseinandergesetzt und eine solche Software entwickelt.

Der Vortrag gibt einen Überblick über verschiedene BCI-Hersteller und -Typen. Das Team hat Techniken aus dem Bereich der künstlichen Intelligenz und genetischen Algorithmen eingesetzt, um die Auswertung von Gehirnströmen aus der Elektroenzephalografie (EEG) zu unterstützen.

Zielpublikum: Entwickler:innen, Projektleiter:innen, Entscheider:innen, Management
Voraussetzungen: IT-Affinität, Grundlagen Künstliche Intelligenz
Schwierigkeitsgrad: Anfänger

Extended Abstract:
Was wäre, wenn ein Brain Computer Interface (BCI) Ihre Gedanken in Befehle zur Steuerung von Software oder Hardware übersetzen könnte? Das TNG Innovation Hacking Team hat sich mit verschiedenen BCI-Geräten auseinandergesetzt und eine solche Software entwickelt. Ein interessantes Gerät in diesem Umfeld ist das OpenBCI, welches Programmierschnittstellen für z.B. C++, Python, Java und R bereitstellt.

Der Vortrag gibt einen Überblick über verschiedene BCI-Hersteller und -Typen, sowie deren unterschiedliche Ansätze zur Erfassung von Strömen. Das Team hat sich dabei der Herausforderung gestellt, Techniken aus der Künstlichen Intelligenz einzusetzen, welche die Auswertung der Gehirnströme aus der Elektroenzephalografie (EEG) unterstützen. Die Sprecher haben verschiedene Experimente und prototypische Showcases unter Anwendung eines OpenBCI implementiert und teilen ihre Erfahrung mit dem Publikum.

In diesem spannenden und unterhaltsamen Vortrag erhalten Sie eine Einführung in die Welt der Gedankensteuerung. Dabei gehen die Sprecher insbesondere auf die Techniken im Bereich des Deep Learning ein, die bei dieser Anwendung zum Einsatz kommen. Die Einführung von z.B. Evolutions-Strategien in Form genetischer Algorithmen zur Optimierung künstlicher, neuronaler Netze runden den Vortrag ab.

Autonomous and automated systems are increasingly being used in IT such as finance, but also transport, medical surgery and industry automation. Yet, the distrust in their reliability is growing. This presentation introduces the validation of autonomous systems. We evaluate in practical situations such as automatic driving and autonomous robots different validation methods. The conclusion: Classic methods are relevant for coverage in defined situations but must be supplemented with cognitive test methods and scenario-based testing.

Target Audience: Testers, Quality assurance, Architects, Requirements Engineers, Product Owners, Software Engineers
Prerequisites: Testing basic know-how
Level: Advanced

Extended Abstract:
Autonomous and automated systems are increasingly being used in IT such as finance, but also transport, medical surgery and industry automation. Yet, the distrust in their reliability is growing. There are many open questions about the validation of autonomous systems: How to define reliability? How to trace back decision making and judge afterwards about it? How to supervise? Or, how to define liability in the event of failure? The underlying algorithms are difficult to understand and thus intransparent. Traditional validations are complex, expensive and therefore expensive. In addition, no repeatable effective coverage for regression strategies for upgrades and updates is available, thus limiting OTA and continuous deployment.

With artificial intelligence and machine learning, we need to satisfy algorithmic transparency. For instance, what are the rule in an obviously not anymore algorithmically tangible neural network to determine who gets a credit or how an autonomous vehicle might react with several hazards at the same time? Classic traceability and regression testing will certainly not work. Rather, future verification and validation methods and tools will include more intelligence based on big data exploits, business intelligence, and their own learning, to learn and improve about software quality in a dynamic way.

Verification and validation depend on many factors. Every organization implements its own methodology and development environment, based on a combination of several of the tools presented in this presentation. It is however relevant to not only deploy tools, but also build the necessary verification and validation competences. Too often we see solid tool chains, but no tangible test strategy. To mitigate these pure human risks, software must increasingly be capable to automatically detect its own defects and failure points.

Various new intelligent validation methods and tools are evolving which can assist in a smart validation of autonomous systems.

This presentation introduces the validation of autonomous systems. We evaluate in practical situations such as automatic driving and autonomous robots different validation methods. The conclusion: Classic methods are relevant for coverage in defined situations but must be supplemented with cognitive test methods and scenario-based testing.