Hoe kunnen we AI inzetten om leerlingen beter te ondersteunen, terwijl de leraar in control blijft? De komende vijf jaar houdt Stephanie Kramer, docent-onderzoeker bij het lectoraat Digital Life (FDMCI) van de Hogeschool van Amsterdam, zich met deze vraag bezig. Zij heeft een NWO Promotiebeurs voor Leraren ontvangen voor het onderzoek ‘Hybride intelligentie in de klas: ontwerpgericht onderzoek naar AI als leerpartner voor hoogbegaafde leerlingen in inclusief basisonderwijs’. Vanwege bezuinigingen van het kabinet zijn nieuwe rondes geschrapt.

Oplossing voor hoogbegaafde thuiszitters

In Stephanie’s onderzoek, dat per 1 september dit jaar van start is gegaan, draait het om hoogbegaafde leerlingen in het basisonderwijs. Zij hebben vaak behoefte aan meer uitdaging en maatwerk, terwijl leraren simpelweg niet altijd de tijd en middelen hebben om dat iedere dag te bieden.

De urgentie is groot: in Nederland zijn naar schatting 70.000 kinderen die niet naar school gaan, en daarvan is 25 tot 40% hoogbegaafd. Passend onderwijs voor deze groep is daarmee niet alleen een uitdaging in de klas, maar ook een maatschappelijk vraagstuk. ‘Kan AI leraren helpen om hoogbegaafde leerlingen de uitdaging te bieden die zij nodig hebben? En hoe doen we dat op een verantwoorde manier?’

Hybride krachten bundelen in een living lab

Het uitgangspunt is hybride intelligentie: menselijke intelligentie versterkt door AI. De leraar, leerling en AI werken samen, ieder vanuit hun eigen kracht — en de leraar houdt de pedagogische regie. Samen met leraren en leerlingen gaat Stephanie onderzoeken hoe zo’n AI-leerpartner collaboratief, adaptief, verantwoord én uitlegbaar ontworpen kan worden.

En dat gebeurt niet alleen achter een bureau. Het onderzoek vindt plaats in en met basisscholen en binnen Digital Life wordt een Living Lab Hybrid Intelligence opgezet. Daar komen leerlingen, leraren, onderzoekers en HvA-studenten samen om ideeën te bedenken, prototypes te bouwen en deze in de praktijk uit te proberen.

Kortom, vijf jaar lang onderzoeken, ontwerpen, testen en bijschaven aan een spannende vraag: hoe zorgen we ervoor dat AI de menselijke intelligentie in het onderwijs versterkt in plaats van vervangt?

Virtuele mensen worden steeds zichtbaarder in het dagelijks leven. Ze lezen het nieuws, begeleiden patiënten, helpen met huiswerk en spelen hoofdrollen in films en games. Dankzij kunstmatige intelligentie en gametechnologie worden ze razendsnel overtuigender en zijn ze overal aanwezig. De vraag is niet meer óf ze ons leven binnenkomen, maar hoe we dat goed regelen.

Daar is de missie van VIVALDI. Als eerste nationale kennisplatform voor virtuele mensen brengt VIVALDI vijf hogescholen samen: Breda University of Applied Sciences, Fontys, Hogeschool Rotterdam, Hogeschool voor de Kunsten Utrecht en Hogeschool van Amsterdam. Samen met partners uit zorg, onderwijs, media en industrie werken negen lectoraten aan de kennis die Nederland nodig heeft.

Dat klinkt als muziek in de oren

Want er liggen fundamentele vragen open. Hoe realistisch moet een virtuele mens zijn, en wanneer slaat dat door? Wat doen ze met ons gevoel van vertrouwen, identiteit en verbinding? Waarom blijven zoveel veelbelovende toepassingen steken als prototype? En wie bewaakt de normen en waarden die in elk virtueel mens worden ingebakken?

VIVALDI onderzoekt deze vragen systematisch, over sectoren en toepassingen heen. Het platform ontwikkelt gedeelde onderzoeksmethoden, praktijkgerichte ontwerpstandaarden en concrete governance-kaders. Bovendien staan het organiseren van een tweejaarlijkse nationale conferentie, samenwerking met beleidsmakers en het delen van de resultaten met de sector op het programma.

Het resultaat: een platform dat Nederland in positie brengt om virtuele mensen niet alleen te begrijpen, maar ook verantwoord te ontwikkelen en in te zetten, nu en in de toekomst.

Expertise vanuit de hoofdstad

De Hogeschool van Amsterdam draagt via de faculteit Digitale Media en Creatieve Industrie bij met twee samenwerkende onderzoeksgroepen, die VIVALDI verankeren in mensgerichte en maatschappelijk ingebedde digitale innovatie.

Vanuit Digital Life is available for download associate lector Child Technology Interaction Simone de Droog betrokken. Zij leidt onderzoek naar de manier waarop sociale robots en digitale personages positief, duurzaam en ethisch kunnen bijdragen aan het welzijn en de ontwikkeling van kinderen in zorg- en onderwijscontexten.

The L.INT-lectoraat Hybrid Realities onder leiding van Esther Hammelburg levert aan dit consortium waardevol onderzoek naar het thema liveness, evenals expertise op het gebied van virtual humans (VH’s) in de context van deathtech. Aangezien dit lectoraat een samenwerking is met het Rathenau Instituut, brengt het bovendien expertise en een netwerk op het gebied van overheid en beleid in, evenals ervaring met de dialoog met burgers en een focus op publieke waarden.

Dark Tech

Students from the Amsterdam University of Applied Sciences’ ICT programme were also present with two interactive robot demonstrations, developed at Studio Dark Tech on behalf of the Digital Life research group. Under the supervision of Professor Somaya Ben Allouch and Associate Professor Simone de Droog, the students explored how social robots can help foster connections between residents of Diemen.

Robot Riddles

For the festival, the students developed, amongst other things, a robot that showcased dance and music from different countries. Visitors were challenged to guess which country the music or dance came from. This sparked conversations about culture, recognition and differences in a playful way. The demo was enthusiastically received by young and old alike.

Daarnaast namen de studenten een robot mee die dieren nadeed in beweging en geluid. Vooral kinderen reageerden hier heel enthousiast op: zij mochten raden welk dier de robot uitbeeldde en gingen zichtbaar nieuwsgierig de interactie aan.

Social earning capacity

The festival demonstrated how technology can stimulate curiosity, interaction and conversation in an accessible way. For many visitors, it was the first time they had actually interacted with a robot. RoboUnity thus served as a fine example of how practice-oriented research and education can contribute to social cohesion.

How the Dream Robot helps imaginative children get through challenging moments

A jab, a tube, or going under anaesthetic: many children find these things nerve-wracking. A new research project is working on a unique solution: the Dream Robot. This robot distracts children through guided imagery. This is a proven method whereby a child can relax through a positive story.

The Dream Robot can move, talk and listen. It looks at children, waves and tells stories in a calm voice. Arine Vlieger, a paediatrician at St. Antonius and an expert in guided imagery, explains: “Fear often plays just as big a role as the procedure itself. If you can reduce that anxiety, the whole experience becomes more bearable. For the child, but also for parents and healthcare providers.”

Guided imagery – also known as medical hypnosis – is less daunting than it sounds. It’s all about focus and suggestion. Arine explains: “When a child imagines something vividly, the body reacts to it. For example: the child listens to a story about swimming underwater and sees all sorts of little fish in their imagination. Then the robot says that sometimes little fish might swim up against your nose. At that very moment, the probe is inserted. The child feels the ‘little fish’ instead of the medical procedure. That’s how the brain and the body work together.”

Guided imagery has been used in hospitals for some time, including by medical educators. However, they cannot always be present during every stressful moment. The Dream Robot project is investigating whether this support can also be provided by a robot.

Three versions of the Dream Robot have been created, for different medical situations: blood tests, tube insertions and anaesthesia. Arine helped develop the stories that the Dream Robot uses. “The child chooses a pleasant setting themselves, such as the beach or a hot-air balloon, and builds a story together with the robot,” explains Arine. “For children undergoing anaesthesia, the child first meets the robot in the day surgery unit. The robot accompanies them to the operating theatre and continues telling the story calmly whilst the anaesthetist puts the mask on. Because the child chooses the story themselves and is actively involved in it, the moment of the procedure is less stressful.”

“We noticed that the children taking part in the research project were completely absorbed by the Dream Robot and fully focused on the story it was telling. Even when someone walked in, they didn’t look up or turn around. Staff and parents also reacted enthusiastically,” says Simone de Droog enthusiastically. Simone is an Associate Lecturer in the Digital Life research group at Amsterdam University of Applied Sciences and the project’s principal investigator. “Parents noticed that their child was relaxing, and that made them feel calmer too.”

The test also highlighted areas for improvement. For example, the robot’s volume was too low, particularly for children who have ear tubes fitted.

The Dream Robot is a collaboration between Amsterdam University of Applied Sciences, Utrecht University of Applied Sciences, St. Antonius Hospital, Wilhelmina Children’s Hospital, Beatrix Children’s Hospital, Erasmus MC and various robotics and hypnosis experts.

The project has been running for two years and is nearing completion. All the findings are now being collated and incorporated into a step-by-step plan for implementation. Following this, a grant will be applied for to ensure that children experience far less anxiety in hospital in the future.

This article has also been published on the website at St. Antonius Hospital.

Een tweede artikel over het project Droomrobot is geplaatst in het vakblad Zorgvisie, waar Simone de Droog (associate lector bij Digital Life) op LinkedIn een mooie reflectie over schreef.

Working together effectively

The project focuses on developing ‘collaborative intelligence’: effective collaboration between doctors and AI in the field of oncology. AI is already capable of quickly and accurately detecting, localising and monitoring tumours based on medical scans. However, this technology is still used only to a limited extent in clinical practice.

That is striking, particularly given that radiologists face high workloads, staff shortages and an increased risk of burnout. Many tasks in oncological imaging, such as preparing radiotherapy plans, are essential, but also repetitive and mentally taxing.

Relevance to everyday practice

A key objective of the project is therefore not only to develop smart AI, but also to design applications that are well suited to everyday practice in hospitals. The focus is on human-AI interaction: how can we ensure that technology actually supports doctors, rather than creating extra work or uncertainty?

The consortium, comprising the Amsterdam University of Applied Sciences (HvA) and the Netherlands Cancer Institute (NKI), is one of 31 research teams to receive funding under the ZonMw Open Competition, from a total budget of over 26 million euros. The competition is known for its fierce competition, which makes the award and high rating of this project all the more remarkable.

The project proposal has been approved by CLICKNL for the next phase, during which CLICKNL will work with the project team to develop a detailed project plan for approval and funding.

A necessary transition

The subject of this project MEDAL is: from medical records to real-time health data: encouraging meaningful daily data collection through design, AI and social engagement. Essentially, the research explores how design, AI and social engagement can motivate people to collect and actively use daily health data. This is crucial for alleviating the pressure on the healthcare system and enabling personalised care.

In recent years, a new approach to participatory healthcare has emerged, in which individuals actively manage their health alongside clinicians, supported by technology and data. In addition to medical records, personal lifestyle data – including experiences, behaviours and lifestyles – are important for this transition, as they enable self-monitoring and improvements in health.

New approaches to participatory healthcare

However, individuals’ motivation to collect data and use it in a meaningful way remains low. Many of the current approaches are technologically advanced but do not fit naturally into people’s daily routines and do not enable them to interpret their data and act on it.

This project demonstrates the power of design and transitions by combining practice-based design research with interventions that accelerate the transition to participatory healthcare.

You should have been there!

The 25th edition was special for several reasons: for the first time in 34 years, the conference took place in the Netherlands, and the number of participants had never been higher: more than 670 registrations from 39 different countries!

Team Digital Life

The AUAS research group Digital Life (Faculty of Digital Media & Creative Industry) was involved in the organisation. Lecturer Somaya Ben Allouch served as Programme Chair Europe and programme manager Monique Schaule Jullens acted as moderator for several parallel workshop sessions.

Using social robots for the elderly and children

In addition, colleague Marianne Bossema, lecturer-researcher at Digital Life, gave a presentation at the conference. She presented her publication “LLM-enhanced Interactions in Human-robot Collaborative Drawing with Older Adults”. This article reports on exploratory, participatory research into the course “Drawing with robots” for older people, which took place in 2024. The course involved collaboration with artists, art teachers and participants to investigate how a robot can support human creativity.

Simone de Droog, associate professor at Digital Life, also took to the stage to share the initial results of the Dream Robot project. Together with Amke Klompmaker, she demonstrated how they combine social robot technology and medical hypnosis to support children during medical procedures.

Meanwhile, the first studies for this project were conducted in a number of hospitals last summer. This phase is expected to be completed by the end of September. Based on the initial insights, the robot is already being modified: think of a more appealing voice, more animation and humour, and improved scripts and procedures to better suit children and healthcare practice.

Finally, PhD candidate Anne Marleen Olthof gave a presentation on Friday about her paper “Attuned Design Practice: Embodied Attunement in the context of HCI, HRI & XAI”, providing a valuable conclusion to this five-day conference.

All in all, it was a wonderful experience to attend the RO-MAN event with a large part of the Digital Life team and to become acquainted with the many interesting visions and research projects in the field of Human Robot Interaction (HRI).

Next year, we will be travelling to Japan!

A really clear definition of this field is still lacking, however. This hinders not only the standardisation of terminology and the reuse of design choices, but also how the wider research community can profile itself as such.

“Human-Artifical Intelligence”

During the annual congress HHAI on “Hybrid Human-Artifical Intelligence”, this year in Pisa (Italy) with some 300 participants, Stephanie Kramer and Michel Oey, lecturer-researchers at Digital Life at the Amsterdam University of Applied Sciences, together with colleagues from Sweden, Denmark as well as from the Vrije Universiteit Amsterdam, conducted a workshop on “Hybrid Intelligence”.

With input from the participants, data were collected to better define the research area of “Hybrid Human-Artificial Intelligence”. In an interactive session, HHAI literature was analysed to identify common disciplinary backgrounds, existing and emerging methodologies, and theoretical frameworks within the community.

In addition, the findings were discussed and interesting transdisciplinary directions for the research field were identified during a session on concept mapping.

Concrete outcomes

The overall outcome of the workshop contributes to an exploratory literature review, including more standardised terminology, an overview of research methods, challenges and future research directions within the HHAI domain. Among other things, this literature review links well to the Digital Life research “Who Takes Care?” in which both Stephanie and Michel have been involved.

Tile wisdom

A key generic message that finally stuck with me came from Stuart Russell, a British computer scientist and one of the keynote speakers to the conference: 'Expectations for AI are very high, it's a huge hype. The consequence that if those expectations are not met, is that we could be in for an “AI winter” in which previous “AI winters” pale into insignificance.“

In short, be wise and, above all, continue to deploy our own human intelligence as well!

Eating and drinking is a headache for many elderly people as well as care professionals. The elderly get less and less appetite and forget to eat. As a result, he or she deteriorates rapidly. But the care professional has too little time to follow this up with everyone. As a result, elderly people can become dehydrated.

'Technology can do all sorts of things, but what adds value? Food and drink, we thought,' says Somaya Ben Allouch, professor at Digital Life. The Intelligent Care Environment consists of sensors and AI that you can put in a home and respond to the specific needs of an elderly person. It does this in a personalised way and provides more support as the older person deteriorates.

Encouraging older people to eat

The moment an elderly person can still get out of bed independently, the lights come on at the chosen time. Moments later, the favourite breakfast is projected on the wall: a plate of oatmeal or a chicken sandwich. In this way, the elderly person is encouraged to eat.

If the sensors detect that the resident does not go to the kitchen, a second signal goes off. This could be a voice saying, 'Don't you want to eat, Mrs Jansen?' Or a smell (e.g. cinnamon or vanilla) or music. If that person still does not go to the kitchen and then skips more meals, a care professional or family member comes to check.

Ben Allouch: 'You can't force people, but you can support them in an empathetic way. Not one size fits all, but according to need and context-sensitive. Not everyone wants three meals a day. And if you have to go to hospital and are not allowed to eat that day, the system doesn't encourage this either.' The stimulation is short-lived each time, otherwise the older person will become overexcited.

We are excited to share the latest updates from ‘WarmStone’, an ongoing design research project of the research group Digital Life, that invites participants to explore how handcrafted, tactile objects can help manage everyday stress. Supported by the Centre of Expertise Creative Innovation, and developed in collaboration with Studio Jiechen and Studio Boey, WarmStone focuses on creating personalized haptic comfort objects through ceramic making.

Project Progress So Far

The first three rounds of workshops have now been successfully completed at Seven and Half Ceramic Studio in Eindhoven, with fantastic participation and engagement:

• Around 30 participants from diverse backgrounds and professions have joined the sessions.
• Over 200 WarmStones have been carefully crafted, each unique in shape, texture, and intent.

At this moment, all crafted pieces are drying naturally, and are being carefully kiln-fired in batches. Once completed, they will be individually packaged and sent back to participants, inviting them to integrate these handmade objects into their daily routines — to carry in a bag, keep on a desk, or use during moments of stress.

What’s Next?

As the first set of WarmStones returns to participants, the next stage of research will begin:

• We will conduct individual follow-up interviews to learn how participants use and experience their ceramic objects in everyday life.
• In June and July, the project will continue with Session 2, where participants will return to the studio to design and create a personalized piece of ceramic tableware (such as a cup or mug) to further explore the possibilities of tactile comfort.

Through these workshops and follow-ups, WarmStone aims to deepen our understanding of how non-digital, handmade objects can offer subtle yet meaningful support for emotional wellbeing.