Author: dsa_admin

April 02, 2025, by Jürg Meierhofer, Expert Group Smart Services

At the ZHAW Service Lunch on April 2nd, 2025, PD Dr. Jochen Wulf and Prof. Dr. Frank Hannich explored the exciting potential of ‘Hybrid Intelligence for Customer Management’, sharing conceptual possibilities and concrete ZHAW project examples on using generative AI across the customer lifecycle.

What happens when human ingenuity meets AI’s power? You get Hybrid Intelligence (HI) – systems designed for humans and AI to collaborate, achieving more than either could alone.

The goal isn’t replacement, but augmentation.

Their ZHAW research explores HI, especially in customer management and technical service. They highlighted powerful applications like personalized content, churn prediction, and augmented customer support.

A key focus is using Large Language Models (LLMs) for tasks like technical assistance, as examplified in their “Guided Maintenance Copilot” project. This involves Retrieval Augmented Generation (RAG) to ground AI responses in specific knowledge.

Developing effective RAG systems requires careful, test-based optimization across factors like search, model selection, and context size. Different models (CPU vs. GPU) and languages show varying performance.

While LLMs can automate tasks from simple translation to complex reasoning, their findings emphasize: there’s no “one-size-fits-all.” Tailored, optimized solutions are key to unlocking the true potential of human-AI collaboration. Ultimately, the session demonstrated that thoughtful design and testing are crucial for successfully implementing hybrid intelligence solutions.

March 11, 2025, by Jürg Meierhofer, Expert Group Smart Services

The IoT Conference 2025, held on March 11 at the Trafo Hallen Baden, was a resounding success, offering an inspiring glimpse into the future of the Internet of Things (IoT). Under the theme “Shaping the Future of IoT – Technologies, Business Models, Security,” around 20 speakers from industry, research and development, as well as providers and users, presented innovative solutions and application examples to the audience of several hundred people.

While the entire Expert Group was invited to attend the conference under special conditions (including some free tickets), the group also actively contributed to the conference in two significant ways. At the start of the conference, Jürg Meierhofer participated in a panel discussion and shared thoughts about value creation models by IoT-based smart services.

• Onboard Rail Monitoring for online noise monitoring for rail head conditioning.
• Structural Health Monitoring for monitoring the structural health of bridges.
• Acceleration of flight tests through advanced data analysis and management platforms.

The presentation demonstrated how Kistler is developing innovative solutions through the integration of IoT and digital products, effectively addressing customer challenges and opening new business fields.

A big thank you goes to the organizers of asut for this inspiring exchange, which made the future of IoT tangible and provided important impulses for the further development and application of these technologies.

Februar 2025, by Klaus Fuchs (Rockstar Recruiting AG), Ivan Sanicola (Innofuse) and Reik Leiterer (data innovation alliance)

On February 13th a webinar about AI-driven tools in the field of scouting,recruiting and matchmaking was organized by the Innovation Booster Artificial Intelligence. This webinar was an inspiring event for the attending recruiters, HR professionals, innovators, and project leaders looking to harness the power of AI and collaborative platforms to drive efficiency, foster innovation, and unlock new opportunities.

According to McKinsey (2025), over 85% of companies will experience skills gaps in key areas in the next few years, with job seekers becoming more demanding and new hires quite volatile, which will exacerbate the challenges. Therefore, there is a need to further develop appropriate digital recruitment strategies while avoiding over-reliance on AI, which could lead to a lack of personalization and thus negatively impact the candidate experience. The challenge in general is to find the right balance between efficiency and personalization and, of course, to operate within the given ethical, legal and regulatory framework.

Two experts from Rockstar Recruiting and Innofuse shared insights, prototypes, and platforms designed to address one of the biggest challenges in today’s professional landscape: connecting the right people with the right opportunities across companies and industries. Klaus Fuchs from Rockstar Recruiting presented an novel AI-based VC screening prototype developed through a Data Booster booster – saving recruiters valuable time by analyzing and pre-filtering hundreds of CVs in parallel – making recruiting faster, cheaper and fairer for everyone involved. 

Ivan Sanicola from Innofuse showcased their innovative platform, designed to connect potential project stakeholders from different organizations through skill-based matching –  enabling seamless collaboration and knowledge sharing.

If you missed this event you can find the hand-outs of the presentation here:

Rockstar Recruiting – Your personal recruitment buddy

Innofuse – The next level of matchmaking

(German version below)

February 12, 2025, by Reik Leiterer, data innovation alliance

The mission of Innosuisse’s Innovation Booster program is to enable radical innovation in particular and to promote the generation of ideas and the development of solutions along the way.

In the field of innovation management, various terms are in circulation that at first glance appear to be easily categorizable and understandable. However, the topic of radical innovation is quite complex in detail, since, among other things, there is a different understanding depending on the industry, culture and size of the company (Tellis et al. 2009). It becomes even more difficult when you want to make the degree of radicality measurable and the radicality of different innovations comparable (Zhang & Tan 2022). This requires quantifiability, which can only be reconciled to a limited extent with the openness of the concept of radical innovation.

This challenge is not new – already Green et al. (1995) stated quite some years ago: “For almost 30 years, innovations have been characterized as radical or incremental. Nevertheless, the construct has not been precisely defined, and ad hoc measures have been the norm in the literature.” And we haven’t really got much further in this area yet – which is why we are at least trying to communicate our understanding transparently below.

What do we understand by the term “radical innovation” in the IB Artificial Intelligence when we evaluate the submitted applications

Our Idea Evaluation Committee consists of external experts and, as part of the evaluation, they assess the technology on the one hand and the market and business model on the other. By technology, we also consider product and process innovations, and a high impact on the market also includes service and social innovation.

The following combinations are thus considered radical:

1) New technology & new market
2) New technology & new business model
3) New technology & high impact on the market

What does this mean in detail? – In principle, we use the well-known 4-field models, such as the one by Henderson & Clark, to assess the radicalness of an idea. These models have – to put it very simply – two dimensions: technology (existing/new) on the one hand and market (existing/new) on the other. Depending on the model, the dimension “business model” or the term “architecture” is also used instead of the dimension “market” and “component knowledge” is used instead of technology. Accordingly, a distinction is made between four different types of innovation in this respect, although there are no clear boundaries, and the transitions are quite fluid:

The challenge here lies in assessing these dimensions, i.e. state of technology and the market/business model. For example, the question, at what point is a technology new? IBM was already working with language models in the 1980s, and in the 1990s, deep learning was increasingly used in the field of NLP… and OpenAI with its chatbot is ultimately part of this series of developments that build on each other. And the same applies to the business model. Business models that have long been established in other areas can be completely new for certain sectors. So, there is room for interpretation, which is why we have decided on a panel of experts who will then decide based on majority rule.

Alright, if you now look at the dimension “impact on the market” instead of the dimension “market” or the “business model”, there is a different diagram:

Can you spot the difference? 😉

From our point of view, the following figure provides a better concept understanding by clearly illustrating the three dimensions (cf. Breakthrough Innovation):

Okay, now we have the categories, but how do you evaluate ideas in terms of which category they belong in? Organizations frequently struggle with measuring the degree and performance of their (radical) innovation activities. Due to the uncertainty and ambiguity involved, current key performance indicators (KPIs) are often not useful in the context of radical innovation, associated with very inconsistent outcomes, and do not appear to measure what they purport to measure (cf. Stiller et al. 2021, Kristiansen & Ritala 2018, Kasmire et al. 2012).

In the evaluation process of the IB Artificial Intelligence, we therefore also consider the following measures, proposed (and described in more detail) by Kristiansen & Ritala (2018).

1) Market orientation: Many technologically novel products fail the actual market test, and firms should therefore pay attention to how they are oriented toward the market for their radical innovation activities.

2) Learning and future opportunities: Firms should encourage opportunities that ensure positive conditions for new opportunities to emerge and even to foster growth in existing business segments. Firms should think about how to de-brief key learnings from the process. Finally, firms should also think about how they would like to use and grow core competencies, including the increase in the relevant knowledge base of the firm.

3) Resource dedication: Radical innovation activities may need substantial intangible resources in the form of highly skilled, cross functional teams. When radical innovation projects mature in the pipeline, they will eventually require an increased allocation of financial resources to get to market. Firms should ex ante decide whether the resources will be available when they are needed.

So much for a brief insight into the theory. If you would prefer a short definition style of explanation, we like the one proposed by Assink (2006) (adapted/extended):

“Radical innovation is a successfully commercialized new product, process, or concept that significantly modifies the demand and needs of an existing market or industry, displacing the previous major players and creating entirely new business practices or markets with significant societal impact. Radical innovation often also requires a corporate culture shift in terms of both internal knowledge management and resource allocation.”

Assink, M. (2006). Inhibitors of disruptive innovation capability: conceptual model. European Journal    of Innovation Management, 9 (12), 215 – 233.

Green, S.G., Gavin, M.B., & Aiman-Smith, L. (1995). Assessing a Multidimensional Measure of Radical Technological Innovation. IEEE Transactions on Engineering Management, 42 (3), pp. 203 – 214.

Kasmire, J., Korhonen, J.M., & Nikolic, I. (2012). How Radical is a Radical Innovation? An Outline for a Computational Approach. Energy Procedia, 20, 346 – 353.

Kristiansen, J.N. & Ritala, P. (2018). Measuring radical innovation project success: typical metrics don’t work. Journal of Business Strategy, 39 (4), pp. 34 – 41.

Stiller, I., van Witteloostuijn, A., & Cambré, B. (2021). Do current radical innovation measures actually measure radical drug innovation? Scientometrics, 126 (2), pp. 1049 – 1078.

Tellis, G.J., Prabhu, J.C., & Chandy, R.K. (2009). Radical Innovation across Nations: The Preeminence of Corporate Culture. Journal of Marketing, 73(1), 3 – 23.

Zhang, J., & Tan, R. (2022). Radical Concept Generation Inspired by Cross-Domain Knowledge. Applied Science, 12, 4929.

Radikale Innovation: Grundlagen

Das Innovation Booster Programm von Innosuisse verfolgt vor allem das Ziel, radikale Innovation zu ermöglichen und eine entsprechende Ideengenerierung und Lösungsfindungen auf dem Weg dorthin zu fördern.

Im Innovationsmanagement sind hierbei verschiedene Begrifflichkeiten im Umlauf, die auf dem ersten Blick gut kategorisierbar und verständlich sind. Im Detail ist die Thematik der radikalen Innovation jedoch recht komplex, da unter anderem sowohl branchenspezifisch und kulturell bedingt als auch abhängig von der Unternehmensgrösse ein anders Verständnis vorliegt (Tellis et al. 2009). Noch schwieriger wird es, wenn man den Grad der Radikalität messbar und die Radikalität verschiedener Innovationen vergleichbar machen möchte. Dies setzt eine Quantifizierbarkeit voraus, was nur bedingt mit der Offenheit der Begrifflichkeit in Einklang zu bringen ist. Diese Herausforderung ist nicht neu – bereits Green et al. (1995) stellten vor einigen Jahren fest: „Seit fast 30 Jahren werden Innovationen als radikal oder inkrementell charakterisiert. Dennoch wurde das Konstrukt nicht genau definiert, und Ad-hoc-Maßnahmen waren in der Literatur die Norm.“  Wir sind auch aktuell in diesem Bereich noch nicht viel weitergekommen – weshalb wir im Folgenden zumindest versuchen, unser Verständnis von radikaler Innovation transparent zu kommunizieren.

Was verstehen wir unter dem Begriff „radikale Innovation“ im IB Artificial Intelligence, wenn wir die eingereichten Anträge evaluieren

Unser Evaluation Committee besteht aus externen Experten und im Rahmen der Evaluation beurteilen sie auf der einen Seite die Technologie und auf der anderen Seite den Markt und das Geschäftsmodell. Unter Technologie verstehen wir auch Produkt- und Prozessinnovationen und der high impact on the market beinhaltet auch Service und Social Innovation.

Dementsprechend gelten folgende Kombinationen als radical:

1. New Technology & new market
2. New Technology & new business model
3. New technology & high impact on the market

Was bedeutet dies nun im Detail? Grundsätzlich orientieren wir uns bei der Beurteilung der Radikalität an dem bekannten 4-Felder-Modellen, wie z.B. das nach Henderson & Clark. Diese Modelle haben – stark vereinfacht erklärt – zwei Dimensionen: die Technologie (existierend/neu) auf der einen und der Markt (existierend/neu) auf der anderen Seite. Je nach Modell wird statt der Dimension «Markt» auch die Dimension «Geschäftsmodell» oder der Begriff «Architecture» verwendet und statt Technologie findet sich «Component Knowledge». Dementsprechend werden diesbezüglich 4 verschiedene Ausprägungen von Innovation unterschieden, wobei es da keine klaren Grenzen gibt, sondern die Übergänge recht fliessend sind:

Die Herausforderung hierbei liegt bei der Beurteilung von Technologie und Markt: ab wann ist eine Technologie neu? Mit Sprachmodelle hat IBM bereits in den 1980er Jahren gearbeitet, in den 1990er wurde zunehmend Deep Learning im Bereich NLP eingesetzt… und OpenAI mit seinem Chatbot steht letztendlich in dieser Reihe von aufeinander aufbauenden Entwicklungen. Und vergleichbar gilt dies auch für das Geschäftsmodell. Für bestimmte Sektoren können Geschäftsmodelle komplett neu sein, die in anderen Bereichen schon längst etabliert sind. Es gibt also Raum für Interpretationen, weshalb wir uns für ein Gremium aus Experten entschieden haben, das dann nach dem Mehrheitsprinzip entscheidet.

Wenn man sich nun statt dem «Markt» oder dem «Geschäftsmodell» sich den «Impakt auf den Markt» anschaut, gibt es ein anderes Diagramm:

Der Unterschied ist nicht so richtig selbsterklärend, oder?

Wir bevorzugen deshalb eher die folgende Abbildung, welche den Impact als dritte Dimension integriert und das Zusammenspiel der verschiedenen Aspekte unserer Meinung nach gut veranschaulicht (vgl. Breakthrough Innovation):

Nun haben wir Kategorien, aber wie bewertet man Ideen im Hinblick darauf, zu welcher Kategorie sie gehören? Organisationen haben häufig Schwierigkeiten, den Grad ihrer (radikalen) Innovationsaktivitäten zu messen. Aufgrund der damit verbundenen Unsicherheit/ Mehrdeutigkeit sind die aktuellen KPIs im Zusammenhang mit radikalen Innovationen oft nicht nützlich, da sie mit uneinheitlichen Ergebnissen verbunden sind und die Aussagekraft bezüglich der Radikalität nicht immer gegeben ist (vgl. Stiller et al. 2021, Kristiansen & Ritala 2018, Kasmire et al. 2012).

Im Evaluierungsprozess des IB Artificial Intelligence berücksichtigen wir daher auch die Aspekte, die von Kristiansen & Ritala (2018) vorgeschlagen wurden.

1) Marktorientierung: Viele technologisch neuen Produkte scheitern am eigentlichen Markttest, und Unternehmen sollten daher darauf achten, wie sie sich bei ihren radikalen Innovationsaktivitäten am Markt orientieren.

2) Lernen und zukünftige Chancen: Unternehmen sollten Möglichkeiten fördern, die positive Bedingungen für die Entstehung neuer Chancen schaffen und sogar das Wachstum in bestehenden Geschäftsfeldern fördern. Unternehmen sollten darüber nachdenken, wie sie die wichtigsten Erkenntnisse aus dem Prozess festhalten können und wie sie Kernkompetenzen nutzen und ausbauen möchten, einschließlich der Erweiterung der relevanten Wissensbasis des Unternehmens.

3) Ressourceneinsatz: Für radikale Innovationsaktivitäten sind möglicherweise umfangreiche immaterielle Ressourcen in Form von hochqualifizierten, funktionsübergreifenden Teams erforderlich. Wenn radikale Innovationsprojekte sich entwickeln, werden oft zusätzliche finanzielle Ressourcen benötigt, um auf den Markt zu kommen. Unternehmen sollten im Voraus planen, ob und wann die benötigten Ressourcen verfügbar sein werden.

Soweit ein kurzer Einblick in die Theorie. Wer aber doch eine Erklärung im Stil einer Definition bevorzugt, dem sei die von Assink, M. (2006) empfohlen (angepasst & erweitert):

„Radikale Innovation ist ein erfolgreich verwertetes, neues Produkt, Verfahren oder Konzept, dass die Nachfrage und den Bedarf eines bestehenden Marktes oder einer bestehenden Branche erheblich verändert, die bisherigen Hauptakteure aus dem Markt drängt und völlig neue Geschäftspraktiken oder Märkte mit erheblichen gesellschaftlichen Auswirkungen schafft. Radikale Innovation setzt häufig auch einen Wandel der Unternehmenskultur voraus, was sowohl das interne Wissensmanagement als auch die Ressourcenverteilung betrifft.“

Assink, M. (2006). Inhibitors of disruptive innovation capability: conceptual model. European Journal    of Innovation Management, 9 (12), 215 – 233.

Green, S.G., Gavin, M.B., & Aiman-Smith, L. (1995). Assessing a Multidimensional Measure of Radical Technological Innovation. IEEE Transactions on Engineering Management, 42 (3), pp. 203 – 214.

Kasmire, J., Korhonen, J.M., & Nikolic, I. (2012). How Radical is a Radical Innovation? An Outline for a Computational Approach. Energy Procedia, 20, 346 – 353.

Kristiansen, J.N. & Ritala, P. (2018). Measuring radical innovation project success: typical metrics don’t work. Journal of Business Strategy, 39 (4), pp. 34 – 41.

Stiller, I., van Witteloostuijn, A., & Cambré, B. (2021). Do current radical innovation measures actually measure radical drug innovation? Scientometrics, 126 (2), pp. 1049 – 1078.

Tellis, G.J., Prabhu, J.C., & Chandy, R.K. (2009). Radical Innovation across Nations: The Preeminence of Corporate Culture. Journal of Marketing, 73(1), 3 – 23.

Zhang, J., & Tan, R. (2022). Radical Concept Generation Inspired by Cross-Domain Knowledge. Applied Science, 12, 4929.

October 01, 2024, by Reik Leiterer, data innovation alliance

On the 29th of August, the Workshop on Challenges in Data Management in Robotics took place in Biel at the premises of the Swiss Cobotics Competence Center (S3C) to bring together robotics enthusiasts, researchers, and industry professionals to discuss and address the pressing issues surrounding data management in robotics applications.

The workshops was organized by the Innovation Booster Robotics, S3C and the Innovation Booster Artificial Intelligence. S3C represents the national hub for industrial and academic partners to jointly develop and test the next generation of industrial cobotic solutions or their components, and to learn how to design and integrate cobotic solutions under human-centric aspects. The Innovation Booster (IB) are Innosuisse supported programs.  IB Robotics is dedicated to fostering radical innovation in the Swiss robotics ecosystem. IB Artificial Intelligence is an open innovation initiative to foster radical innovation using socially- and economically viable AI.

The aim of the full-day workshop was to use keynotes, case studies and subsequent breakout sessions to provide participants with insights into data-related challenges in robotics and to jointly develop possible solutions for overcoming them.

Four complex subject areas with their respective challenges were defined:

• Challenge 1: Co-Working and Learning
  
  Understanding human practices within the context of human-robot interaction is crucial for robots to adjust their behaviours appropriately. Not only do technical tasks and ethical considerations come into play, but also the fact that different individuals interact in unique ways, making the development of universally applicable algorithms extremely difficult. Furthermore, ensuring precise communication and task allocation in the human-robot workflow is essential, especially considering the proximity between humans and robots.

• Challenge 2: Data Management
  
  Most robots work with both filtered and uncensored data, not only for robot training and learning tasks but also directly acquired in-process by physical sensors. Nowadays, tasks such as training may require a large amount of data. This can be challenging in terms of data overload, costs of data analytics integration, and robust data management practices.

• Challenge 3: Responsible Collaboration
  
  Both humans and machines occasionally fail at specific tasks, making it essential to consider failure scenarios in developing robust interaction frameworks. By understanding and leveraging failure data, we aim to improve the design, development, control, and robustness robots, ensuring more resilient and adaptive human-robot interactions. There is a need to explore methods for identifying and capturing failure events, analysing their causes, and utilizing this data to improve the reliability and performance of robotic systems.

• Challenge 4: Self-improvement of Robots
  
  One aspect of robustness involves the capacity for a robot to enhance its knowledge and behaviour autonomously. This means it should possess a degree of flexibility to adapt and apply abilities to new situations as needed. Achieving this context-driven, adaptive autonomy, which relies on common-sense knowledge and practical manipulation tasks, demands extensive programming and often involves on-platform data management & analytics.

The topic “Coworking and Learning” was opened by Baptiste Busch – CEO and Cofounder of AICA. He pointed out the demand for advanced and flexible robotic solutions in the reindustrialization process and the challenges around the sensing of and co-working with robots. Opportunities in this regard were seen around sensing-based control, behavioural programming increasing self-adaptation capabilities, and the use of foundation models. On the way towards cageless cohabitation, changes in the collaboration between robot manufacturers, integrators and end-users would be necessary while regulatory changes will bring additional challenges (e.g. ISO 10218-1:2011 and EU AI Act).

Björn Jensen from HSLU and Yves Albers-Schoenberg of Roboto AI addressed the challenges around “Data Management” – from data retrieval and data analysis to data augmentation and the generalizability of data models. It was discussed, how the combination of data, models, and simulations is the base towards data-centric robotics – and with reference to the Moravec’s paradox, that assembling the right data and deploying the right models for sensory processing and perception skills require enormous computational resources.

After the panel discussion, the participants were assigned to the two challenges and worked together on possible solutions and what would be the requirement for starting projects on the identified solutions.

The outcomes of the presentations and discussions around the “Co-Working and Learning” challenge were:

• Human-robot interaction is constantly evolving, and collaboration requires robots to manage the variability in human behaviour, which is not easily predictable. Current approaches rely heavily on data, but there is also a need to focus more on understanding human behaviour beyond scripted interactions. Empowering collaboration requires better data retrieval, analysis, and generalization techniques.
• Reinforcement learning in simulated environments helps robots become robust, but translating this learning to real-world environments remains difficult. Robots often struggle to adapt outside of simulations, indicating a need for them to learn more effectively from human behaviours.

Based on the identified challenges, the following ideas were developed and possible collaborations between the participants discussed:

• Develop a more effective system for categorizing the skills required based on specific tasks and types of collaboration. And create a unified platform to hare categories.
• Establish industry-wide safety standards for human-robot collaboration that are universally applicable, including the creation of adaptable standards, norms, and guidelines for various environments.
  
• Promote partnerships between industry and academic institutions to utilize common datasets on skills, bridging the gap between research and practical applications. This approach could help accelerate the transition of innovative solutions from the lab to the factory floor.

The outcomes of the presentations and discussions around the “Data Management” challenge were:

The following ideas were developed:

• Consider ways to handle the variability in human behaviour. These could be trained on diverse datasets to improve robots’ adaptability in real-world situations, allowing for more seamless and effective collaboration.
• Create a centralized platform for managing large-scale robotics data. This platform could provide standardized tools for data collection, curation, and processing, making it easier to develop and refine foundation models.
• Establish open-source repositories for robotics datasets, curated with respect of the real world.
• Develop new data compression methods to handle the vast amounts of data generated by robots. This would enable more efficient storage and faster processing, making it easier to utilize data for training and collaboration.

After the intensive and inspiring morning, there was a networking lunch with the opportunity to have a look on the various services and test options offered by the S3C, including i) realistic, industrial-like test environments with confidential testing conditions, ii) adaptable, modular test benches and setups, and iii) scalable options from quick tests to extensive, long-term testing.

After lunch, the afternoon session starts with the topic “Responsible Collaboration”. Andreas Hufschmid from PILZ explained why a holistic safety concept is required in the field of industrial robot systems and while dealing with HRI. This considers both machinery safety and industrial security and is mainly addressed in the ISO 10218 standard, created in recognition of the particular hazards that are presented by industrial robots and industrial robot systems. For this, risk assessments should be conducted to determine what the protective measures should be.

Andrei Cramariuc from the Robotic System Lab (ETHZ) presented the challenges in cases, were robots need to respond to human interactions or should perform complex, everyday tasks in dynamic environments. For this, approaches such as kinaesthetic guidance or teleoperation are currently explored.

After the break-out sessions again, the outcomes around the “Responsible Collaboration” challengewere presented:

• Safety is a critical concern in human-robot collaboration, with collision forces being a primary issue. Current safety standards are essential but require ongoing verification and validation to ensure effectiveness as regulations evolve.
• Robots must be able to respond to human interactions and perform complex, everyday tasks in dynamic environments, i.e. dealing also with uncertainties and adapt to changes. Different approaches (e.g. kinaesthetic guidance and teleoperation) are being explored to improve how robots learn and interact in real-time scenarios.
• Balancing safety with innovation is a major challenge, especially in public spaces and assistive robotics. There is a growing recognition that safety by design must be integrated into the development process, with reinforcement learning playing a potential role in achieving this balance.

Related to the “Self-Improvement of Robots” the following challenges were identified:

• Robots often struggle when transitioning to real-world applications. The challenge is to enable robots to adapt and improve their performance based on real-world experiences, learning from human behaviour.
• Research is focused on enhancing robots’ capabilities beyond basic tasks, such as using their entire body for interactions. Efforts are underway to develop more sophisticated data structures and learning strategies, like self-learning with tensor networks, to improve robots’ adaptability and performance.
• A key area of focus is training robots to specialize in specific tasks, such as assembly, by refining foundation models. The ability to standardize data and share solutions across platforms could significantly advance the self-improvement of robots.

Based on these challenges, the following ideas were discussed and shaped:

• Facilitate the sharing of models and simulations across different platforms, including sharing of successful learning strategies among different robots and platforms (potentially through a centralized knowledge base).
• Extend the application of reinforcement learning beyond simulations by enabling robots to learn directly from real-world interactions, possibly within controlled environments where they can safely experiment and refine their skills, i.e. gradually expanding their ability to adapt to real-world norms as they become more capable. In addition, implementing real-time feedback loops could be of interest to allow robots to instantly adjust their actions based on human responses. This approach could also encourage mutual learning between humans and robots.
• Develop specialized foundation models tailored to specific tasks, such as assembly or inspection, to enhance efficiency and effectiveness in improving robot performance within these domains.

Ideas discussed during the workshop can now be submitted to the IB Robotics (next DL 13th of September) or to the IB Artificial Intelligence for further funding in order to develop them further in terms of feasibility, viability and desirability.

September 08, 2024, by Reik Leiterer, data innovation alliance

The aim and motivation of the RiskON initiative are to promote the further development of risk management in Switzerland under the patronage of sminds AG/N9 House of Innovation, the program management of the University of Zurich, especially of the UZH Innovation Hub, including business partners such as Bank Julius Baer, LGT Bank Switzerland, Pictet, EFG International and the Association of Swiss Cantona Banks.

Actively joined strategic partners in 2024 were Swiss Financial Innovation Desk (FIND), Swiss Risk Association and the Innosuisse Innovation Booster Artificial Intelligence. The latter will allow for the practice transfer after the RiskON Hackthon – a practice transfer for co-innovation between banking representatives and students, guided by the UZH faculty.

The RiskON 2024 had three concrete challenges with 15 teams and 53 students to work on and produced use cases and approaches for the banks involved:

• Challenge-1 ‘Use of new technologies for risk events classification and analysis’
  Financial institutions, as part of their risk frameworks, are required to capture, investigate/analyze and report on risk events. Given the volume of such events, their handling and analysis may be cumbersome and resource intensive. So how to leverage new technologies to facilitate the classification of risk events and their analysis?
• Challenge-2 ‘Fraud Risk Detection – How artificial intelligence can help’
  Fraud risk, external and internal, is currently among the top risks in the financial sector. Big data and the rapid development of artificial intelligence has, unfortunately, also contributed to that over the past few years. On the other hand, AI-methods can be efficiently used by banks for early detection of fraud risk. In this context, a pragmatic approach for using AI methods and tools as part of a fraud risk detection tool with a focus on Private Banking business profiles is required.
• Challenge-3 ‘How can AI be leveraged and enabled to enable CRO employees to work more efficiently’
  By focusing on either completing counterparty due diligence questionnaires, enhancing the Anti-Money Laundering (AML) alert resolution process or streamlining, harmonizing, translating, and updating legal client forms using advanced analytics techniques, Optical Character Recognition (OCR), and NLP, ideas for improving, scaling and automating tasks performed by Chief Risk Officer (CRO) employees are needed (incl. considering the associated risks, limitations and implementation cost).

For the winning teams, these steps are clear: the innovative solutions will receive funding and support to ensure practical implementation. Within six months, a Minimum Viable Product (MVP) or Proof of Concept (POC), showcasing the feasibility, viability and desirability of the ideas, will be put into motion. The innovation teams will also get the opportunity to co-author articles and white papers with UZH professors, gaining access to cutting-edge research, enhancing academic credentials and thereby allowing for academic transfer.

July 4, 2024, by Reik Leiterer, data innovation alliance

The Swiss Conference on Data Science (SDS) is Switzerland’s premier event for applied data science. The conference brings together leaders and science and business experts to exchange ideas and drive innovation in products and services, with a focus on the Swiss market. The SDS2024 took place in Zurich on May 30-31 at The Circle Convention Centre, Zurich Airport. If you want to get an impression how it was, have a look on the SDS2024 Flashback Video!

The 2-day conference started with an interactive workshop day to provide in-depth, practical and application-oriented insights into the latest developments in the field of data science and Artificial Intelligence. Over 450 participants took advantage of these opportunities and were able to benefit from exciting and professionally prepared and conducted workshops. Two of the workshops were supported by the Innovation Booster Artificial Intelligence to identify challenges and discuss possible ideas for radical solutions.

Next-Gen Cleantech Solutions: Mining Insights from Media and Patent Data with Natural Language Processing (NLP) and Large Language Models (LLMs)

The first workshop addressed challenges and solutions in the cleantech sector and was organized and moderated by HSLU (Guang Lu), ETHZ (Susie Xi Rao), FHNW (Daniel Perruchoud) and Equintel GmbH (Janna Lipenkova). All slides, data and further information are provided here.

The workshop started with a talk about Disentangling the Global Innovation Landscape by J. Lipenkova and expert techniques were presented for analysing patent and media data for cleantech innovation including NLP, LLMs, RAG, and LLMs-augmented recommender systems. These inputs were then applied in a hands-on session and transferred to specific use cases, where the participants were able to try out the presented LLM-powered cleantech question-answering and recommendation system.

Generative AI for Well-Being

The 2^nd workshop deals with the potentials of the new technologies around generative AI to help people to fight stress and increase their well-being. This workshop was organized by BFH (Souhir Ben Souissi, Mascha Kurpicz-Briki, Yannis Schmutz, Tetiana Kravchenko, Christoph Golz).

The day was concluded with a networking Apero and the presentation of the Swiss Viz Awards.