Artificial intelligence and UX

Our vision is to make technical products and innovative technologies in industry user-friendly and appealing. We strive to create working environments that take people's needs into account, create an enriching atmosphere and promote people's health. This also includes new technologies such as artificial intelligence (AI). Here, too, we focus on sustainably increasing well-being at work. We are committed to reducing the often complex world of artificial intelligence to the essentials and providing users with comprehensive support along the way. Explainable AI (XAI) plays a major role here in opening up the black box of AI and making the underlying processes understandable.

Closing the gap between AI and humans in industry

Does this sound familiar? You collect data, try to derive benefits from it and perhaps even have a proof of concept. But no one is using your AI solution yet? Decisions are still based on gut feeling, as users are reluctant to use AI or machine learning (ML) systems. There's a lack of trust and users may even be afraid to use the systems. This is not an ideal situation. Especially in times of a shortage of skilled workers, it is important to offer attractive jobs in industry. OpenAI has shown how important an optimal user interface is to make people want to interact with AI.

AI for people: Explainable AI and UX

Explainable AI (XAI) significantly contributes to a positive user experience by making the complex functions of AI and ML systems transparent. Decisions users make are supported, for example, by displaying the accuracy of predictions and showing the decision-making paths of AI models.

When users understand which criteria an AI uses to make recommendations or adjust system settings, this does not only promote understanding of the technology, but it also builds trust and acceptance. An effective human-machine interface ensures outstanding user experience, and ideally also confidence, operating safety and enthusiasm. It integrates users' needs and expectations.

A human-machine interface of this kind can help visualise the complex interrelations of AI decisions in a user-friendly way. It also contributes to the interaction with the AI or ML systems in such a way that the human using the machine experiences pleasant and welcome assistance without giving up control.

Legislation and standards are forcing the industry to act
The increasing adaptation of standards to the topic of artificial intelligence (AI) is particularly evident in the EU's recently adopted AI Act, which was passed at the end of 2023 and will come into effect shortly. This once again underlines the enormous importance of Explainable AI. It is no longer just a recommendation, but rather a binding measure. This development clearly shows that transparency in decision-making processes is not only necessary in high-risk areas.

Our UX approach for Explainable AI

We have been working intensively on the topic of AI for years. Our experience to date has shown that Explainable AI is a significant approach to UX. Our aim is to strengthen the connection between people and AI in order to use this technology positively for even more user-friendly hardware and software. With Explainable AI, as with our approach to designing an Overall User Experience, we want to increase wellbeing at work. We recognise that it is crucial to reduce complexity to a minimum and to take users on this journey with us.

Our goals

Build understanding and trust among users
Transparent presentation of AI recommendations and predictions
Communicating the benefits for people in order to make AI understandable as a valuable assistance system
Establish AI as an assistance system for people and make it attractive
Transparency thanks to recognizable AI, i.e. interactions in which users explicitly recognize the presence or availability of AI

How your product will benefit from this

Greater acceptance and success of your software among users
Promotion of collaboration between AI and humans to provide the best possible support for people and, of course, to improve the performance of the machine
Improved operating safety
Promotion of successful strategies and parameter sets for more efficient machine performance (e.g. speed, quality, material and energy consumption)
Increasing overall equipment efficiency (OEE)
Increasing the efficiency of the AI application
Promotion of innovations

We have developed solutions for:

Integration into the HMI context without losing focus on the actual HMI
Presentation of the confidence level to clarify the degree of certainty of AI statements and create trust
Patterns for assessing data quality
Visualisation patterns of model interpretations, e.g. showing the contribution of certain features (e.g. pixel values, word frequencies) to model predictions
Patterns for parameter settings based on historical data or settings
Patterns that visualise how parameter changes affect the OEE
First steps on the way to fulfilling documentation obligations by displaying e.g. confidence levels
...

Our services
We would like to share our ideas, experience and knowledge on the topic of AI, especially Explainable AI, with you and offer various formats. Our goal is to ensure that your software is accepted by users and thus becomes successful, not despite AI but because of it. Through our services, we discover new innovation potential and show how AI can improve your product. We offer the following methods, among others:

AI user research
AI user journey workshop
XAI UI patterns (e.g. visualisations of AI models or AI results, controls for decisions based on comparisons)

Strong together
We are not data analysts, data scientists or AI experts, but specialists in the field of user experience (UX). As such, we form a powerful team with your experts. In addition, we continuously maintain and intensify the exchange with experts from our extended network.

AI and user interfaces: How the use of AI can change forms of interaction

The integration of artificial intelligence is fundamentally changing the user interfaces of technical products and machines. Conventional interaction patterns are increasingly giving way to innovative approaches and UIs are becoming more and more flexible. At the same time, the need for explainability and monitoring is growing in order to enable well-founded assessments and decisions. Chatbots are only part of the solution in very specific situations.

Interaction patterns: chatbots and assistants
Thanks to AI, it is possible to respond to users in real time, allowing requests and needs to be met immediately. In particular, this includes chatbots that can answer complex questions and virtual assistants that take over tasks. These efficient tools often lead to drastic changes on the interface side.

In the past, chatbots were often used excessively in web design due to immature technology and had the reputation of being a poor substitute for human service tasks. Thanks to AI, this has changed. The use of chatbots is becoming more attractive as they can now not only answer complex questions, but also respond in seconds. Despite these advances, however, we should keep in mind that chatbots can occasionally "hallucinate" due to AI technologies by generating inaccurate or misleading information. A conscious and critical approach is therefore important.

Interaction methods: voice control, gesture control
Advanced interaction methods are being further developed thanks to AI. The performance of the systems enables users to communicate directly with the software. The quality of voice output and content has become so good that it feels as if you are talking to a human counterpart. The accessibility of this interaction method opens up enormous possibilities compared to keyboard and mouse operation or touch operation.

This is where the journey will take us: Personalisation, real-time customisation and recommendations
Personalised user interfaces have established themselves as the gold standard. Thanks to AI, they can now also adapt to individual preferences in real time, be it in terms of colours, fonts or functions - there are no limits to creativity here. This depends on the characteristics and skills learned by the AI as well as users' expectations of the system.

In AI-supported user interfaces, recommendation interactions, which are already a familiar concept from the world of social media, become essential patterns in order to continuously improve the AI. At the same time, AI offers users predictive recommendations. Instead of just waiting for user instructions, it anticipates needs and delivers contextual content.

Not only real-time recommendations are possible. AI also enables predictive planning of activities by anticipating future needs and requirements of machines and users and making proactive suggestions.

AI and user experience in a nutshell
Artificial Intelligence (AI) has a significant impact on the UX of systems. In a world where AI is becoming more and more integrated into everyday life, it is crucial that users can understand how AI systems make decisions. Explainable AI (XAI) makes it possible to open up the black box of AI and make the underlying processes understandable. UX design closes the gap between data science and users' needs. Good UX design ensures that information is accessible and presented in a clear and user-friendly way to increase trust in AI and ML applications. A user interface can make complex AI models and their decision-making processes transparent and understandable for users.