Præsentationer 2020

Du kan læse mere præsentationerne på årets konference på denne side, som løbende opdateres.
Det endelige program offentliggøres senere og der kan forekomme ændringer.
Du kan tilmelde dig årets konference her: TILMELDING
Tilbage til Konference 2020
Read more about the conference presentation below which we update continuously.
The final program will be published later and changes may occur.
You can sign up for this year’s conference here: SIGN UP
Back to Conference 2020


Experiential Learning Testing AI: Future Proofing Your Role as a Software Tester (in English)

In this workshop you will gain hands-on experience of testing AI. In doing so, we will explore the effectiveness of current software testing methods, how they may need to be adapted for testing AI, and look at new approaches developed for testing AI.
With AI, software is moving from deterministic outcomes where correct behaviour is relatively clear and well-known, to outcomes that are not necessarily understood, even by those designing and developing systems. AI is often treated as a black box, because its complexity makes its behaviour so hard to explain. In addition, through its ability to both scale and rapidly change through machine learning, AI software has the potential to affect more people in unpredictable ways.
This adds additional complexity and responsibility to testing, and we will look at the importance of exploratory testing as a method to understand an AI system, and help work with AI developers to improve the quality and the ethics of their systems. As we test, we will assess the efficacy of our testing with regard to key issues in testing AI such as: bias; validity and accuracy; transparency; and unpredictability. We will explore techniques for probing AI models, revealing and explaining the patterns inside the algorithms, and how these relate to existing test methods.
We will also do a deep dive into several AI application scenarios, and consider how to devise testing methods that would address the risks of unpredictability in machine learning systems, especially when drawing on live data sources like social media and text chat.
Throughout our testing, we will reflect on the changing skill set and job of the software tester of the future, and consider the skills that testers will need to enable them to stay at the front of their profession as new software techniques become an increasingly significant part of the world of software development.
Navn: Chris McKillop, Virksomhed: Turalt
Workshop is in English.
Tilbage til toppen.

Mobile Test Automation (in English)

Mobile devices have spread rapidly around the globe with an estimated number of more than 5.1 billion people connected to mobile services by the end of 2018 and expected to reach 5.8 billion by 2025 (71% of the world’s population) with 5 billion using mobile internet.
From these, at least 60% (79% in 2025) are estimated to be connected using smartphones, and frequently, with more than one device. With these numbers, and even with the smartphones market dominated by two main platforms, Android and iOS, for the Software Testers, the number of different combinations between device models and operating system versions constitutes a considerable challenge requesting techniques like 80/20 and risk-based prioritization.
Additionally, due to the nature of the market and the need for shorter release cycles, most (if not all) of the mobile development projects are following agile methodologies.
This creates a context where test automation is no longer optional or just a nice to have. It’s a mandatory requirement for the test teams no matter if it is mobile web or mobile app development.
In this introductory workshop besides going through the most common tools for mobile test automation and diving in with an hands-on exercise (both mobile web and mobile app), we will also review the main concepts (including a mobile test strategy) and success (or fail) factors for a test automation project.

Who should attend this workshop?
• Manual testers that want to do mobile test automation

• Tester experience.
• Windows laptop with permissions to install software.

Workshop Content
1. Test Concepts (refresh)
2. Test Automation (five Ws and one H)
3. Mobile Test Strategy
4. Required Skills
5. Tools (types and selection)
6. Hands-on Mobile Test Automation
a. Selenium
b. Appium
c. Robot Framework
7. Recommendations and Takeaways
Navn: Joel Oliveira, Virksomhed: PSTQB – Portuguese Software Testing and Qualifications Board
Workshop is in English.
Tilbage til toppen.

Præsentationer (Presentations)

AI and the Future of Software Testing (in English)

What is AI? What does it mean for software testers? Is AI and automation coming for your jobs? How can software testers positively impact the future of AI?
There is a lot of confusion about what AI is, and what it isn’t. We are increasingly seeing its impact on individuals and society – its potential for good, and its potential for harm. Using case studies, we look at AI’s big failures and identify what led to that failure. These failures raise a number of key challenges that will be faced by software testers in the new age of AI:
• Bias: societal bias is often baked into AI systems, and we will discuss efforts to produce tests which both highlight and mitigate this risk.
• Validity and accuracy: accuracy of the model is often the prime focus in testing, but we will show that the validity of the model needs insightful testing.
• Transparency: We look at challenges in testing AI which is often a black box.
• Unpredictability: we are still learning how people use AI systems, and often that is not in a way we could predict. How can we test the unpredictable?
Whilst it may be a fear that AI and automation will eventually replace the role of software testers, we will learn about the shortcomings of using AI and automation in testing, and show that, in testing, the human should always have oversight.
Navn: Chris McKillop, Virksomhed: Turalt
Presentation is in English.
Tilbage til toppen.

Aktivistisk testrapportering – gør en forskel (på dansk)

Testrapportering er en vigtig del af testmanagerens opgaver, men det er ikke altid lige let at komme ud over kanten med det ønskede budskab. Testrapporteringen skal vise det sande billede af situationen, og modtages ikke altid med applaus af modtagerne. Som testmanager kan man nemt blive presset ud i en situation, hvor andre gerne vil have omformuleret konklusioner m.m. Hvordan håndteres dette?
Navn: Ole Chr. Hansen, Virksomhed: Q Nation A/S
Præsentation er på dansk.
Tilbage til toppen.

Encouraging Children into Software Testing (in English)

Since everything is being digitalised society needs IT and testing skills like never before. There is a persistent lack of both programming and testing skills on the job market.
More children learn how to code and create programs but not how to make sure the programs work well by testing the programs. For some children software testing may be their best and most viable way into the fascinating world of information technology.
Regardless if children code, test or just use software they should have solid software testing skills. Such skills enable them to look at the software they use and understand its pros and especially its cons. It all starts with awareness and understanding of software testing.
In order to teach children how to do software testing it is important to understand how children learn through playing, remembering, identifying with, and exploring. Kari has created one solution to stimulate children to learn about software testing: a fantasy book about software testing for children.
In this presentation Kari Kakkonen explains why we should encourage children to start with software testing at an early age. He shares some examples of how fantasy can represent software testing:
• Dragons are defects
• Knights are developers and testers
Kari discusses different defects people have experienced and transform them into dragons. All attendees get to try it by drawing a dragon/defect! In addition Kari shares some extracts of the book to everyone and if the book is published he will give away a few copies to the most active participants.
Navn: Kari Kakkonen, Virksomhed: Dragons Out Oy
Presentation is in English.
Tilbage til toppen.

How We Deal with Large Scale Test Automation Challenges at Unity (in English)

Test automation is completely indispensable when you need to deliver and support multiple versions of a complex software product running on more than 25 platforms. And its challenges get to a completely new level when you run thousands of tests millions of times a day.
What pitfalls are waiting for you on such a complex setup and how to approach them?
Should you consider these problems from the very first tests you code or solve issues one at a time as you grow?
In this presentation Tetiana shares her experience of solving large scale test automation challenges in the R&D team in Unity Technologies that builds a world-leading game engine.
Tetianan will cover some of the main problems causing long feedback time and flaky tests:
• Slow tests
• Better quality and usability of test tools
• Improve developers awareness of the quality of the tests
Navn: Tetiana Krylova, Virksomhed: Unity Technologies
Presentation is in English.
Tilbage til toppen.

Rise of the Machines – AI to the Test (på dansk)

“At 02:14 am Eastern Time on August 29, 1997, Skynet becomes self-aware” – dette er et af de legendariske citater fra Terminator filmene.
Terminator 1 blev vist første gang I oktober 1984, med en dommedagsvision om hvad der ville ske godt og vel et årti ude i fremtiden; Skynet, et forsvarssystem baseret på kunstig gruppebevidsthed og kunstig superintelligens, begynder at opfatte menneskeheden som en fjende og bliver egenrådig.
I filmene går det galt, rigtigt galt. Vi får aldrig rigtigt forklaringen på hvorfor, for det var, jf. den ansvarlige, General Robert Brewster, gennemtestet… – tydeligvis ikke nok, eller ikke på den rigtige måde?
Ligesom med mange andre teknologier, kræver det en særlig indsigt at kvalitetssikre AI. Selvfølgelig skal testen forholde sig til forretningsværdi og risiko. Men testen er i høj grad også en kontrol af sandsynligheder og forudsigelser, der tilmed kan foregå på et non-deterministisk grundlag. Testerens indsigt i og forvaltning af data er derfor vigtigere end nogensinde. Det samme er forståelsen for teknologien, dens anvendelse, styrker og svagheder.
Den test man ikke tog alvorligt hos Skynet, står mange virksomheder nu overfor i deres arbejde med at implementere AI modeller. Er kvaliteten rigtig er der store forretningsmæssige gevinster at hente. Så meget desto større er potentialet for katastrofe, hvis det går galt.
I dette indlæg vil vi forsøge at kaste lidt lys over hvor AI og ML er i dag, hvilke udfordringer, der er med at teste kunstig intelligens, hvordan man griber testen an, og hvordan vores fremtid bliver sammen med robotterne.
Navn: Helge Nymand, Virksomhed: Testhuset A/S
Præsentation er på dansk.
Tilbage til toppen.

The Usage of Test Design Techniques in the “Real” World (in English)

We all know from ISTQBs principle number two (Exhaustive testing is impossible) that we cannot test everything. This is just one of the arguments for using test design techniques when designing test cases.
When I have finished an ISTQB course I often leave participants that is excited about coming home into their organization and start working with the test design techniques. If we then meet up months later, and I ask them how it is going with incorporating the techniques in their analysis and design, the feedback I get is that the techniques is not being used or only a few is being used.
Why does this happen so often. The reasons are many, some have difficulties in choosing the appropriate technique, some have requirements that does not fit perfectly into a technique and some feel that the analysis and design will be more time consuming when using the techniques. How do we get to use the test design techniques in a world where the requirements are not written to match a specific technique?
I will by using concrete examples from the “real” world answer this question along with the other difficulties we as testers meets in regards of choosing the appropriate technique. I will show how to choose your test design technique based on the risk level, test type, type of test basis and the concerned test level. I will work with requirements written in different ways and with choose that has not been specified.
Navn: Annemette Clement, Virksomhed: Capgemini
Presentation is in English.
Tilbage til toppen.

TQED – a Simple Test Design Technique (in English)

Adam Roman has developed a simple and universal model for test design, called the TQED model. It may be viewed as a fault model, but also as a handy tool for boosting the tester’s creativity when designing the tests or performing static testing.
According to Adam, we have somehow lost the ability to reflect deeper on the foundations of our testing craft. And if we manage to analyse things in depth we often struggle to transform this reflection into a working solution. Instead we tend to rely on a new tool or sophisticated technology to solve our problems. But thinking can in many cases be much more effective approach.
Following the conference’s motto, Adam will begin from the theoretical perspective and explain the origins and justification of the model (inspired by physics), its structure and components. TQED is also compared with other well-known test design techniques. Then, Adam will present some practical applications of this technique in generating the test ideas and also present a scientific proof that the model really works.
Navn: Adam Roman, Virksomhed: Jagiellonian University
Presentation is in English.
Tilbage til toppen.

DSTB på LinkedIn


Du kan tilmelde dig den åbne gruppe her

DSTB på Twitter

Du kan følge os på @dstb_dk og konferencen på #DSTBKonf.Twitter_logo_blue

Digital Quality and Testing Meetup

Are you interested in quality and testing? Then you can meet like-minded people in our group on Meetup called: Digital Quality and Testing Meetup in Denmark.

Click the link to join the group:

You can read more about the group here: