Meet Lola Jo Ackermann and Alexandre de Montleau at Mensch und Computer 2026 and explore two interactive CDHI demos that make digital health interventions more adaptive, transparent, and responsive.

Digital health interventions face two deceptively simple questions: What should happen when engagement begins to decline? And when is a person likely to be receptive to support? At the Centre for Digital Health Interventions, two new interactive demonstrators approach these questions from complementary directions. From 30 August to 2 September 2026, we will present both projects at Mensch und Computer 2026 at the University of Duisburg-Essen in Duisburg, Germany.

ADHERE: Designing for non-use

Most digital health demonstrations focus on what happens when people engage. ADHERE deliberately asks what should happen when they do not.

Developed by Gianluca Piol, Lola Jo Ackermann, Marcia Nißen, and Tobias Kowatsch, ADHERE-MCP is a reusable service for predicting and responding to emerging non-adherence in mobile data-collection and intervention studies. Before a scheduled interaction, the service reconstructs the participant’s recent history, estimates the risk of missing the next touchpoint, explains which factors contributed to that estimate, selects a preventive strategy when appropriate, and logs the decision for later inspection.

The demonstrator integrates ADHERE-MCP with ScrollWise, a five-day chatbot-delivered micro-coaching intervention addressing doomscrolling and problematic social-media use. Visitors can explore different scenarios, ranging from consistent engagement to early warning signs and repeated missed touchpoints, and observe how the system’s risk assessment and response change.

As shown in the demo video, an additional message grounded in an established behaviour-change technique can be scheduled shortly before the next planned session. The goal is to make support proactive, explainable, and reusable across different studies, not simply to send more reminders after an interaction has already been missed. Lola Jo Ackermann will present the ADHERE demonstrator in Duisburg.

PULSE: Finding a more receptive moment

Even valuable support can be ignored or experienced as disruptive when it arrives at the wrong time. PULSE, the Personalized User-Level Scheduling Engine, makes the timing decisions behind just-in-time adaptive interventions visible and configurable.

Developed by Alexandre de Montleau, Marcia Nißen, and Tobias Kowatsch, PULSE is demonstrated together with ScrollWise-JITAI. When the intervention identifies a critical period, for example, a period in which someone may be at risk of doomscrolling, it asks PULSE to recommend a promising time for support within that window.

PULSE addresses the cold-start problem through progressive personalization. It can begin with population-level interaction patterns, move to evidence from a cohort of similar users, and eventually use the individual’s own interaction history as more data become available. Its response includes a recommended time, an estimated interaction likelihood, the evidence level used, and a rationale.

The video walkthrough follows a first-time user for whom no personal interaction history is yet available. PULSE therefore draws on a tailored cohort, recommends 7:00 p.m., and ScrollWise-JITAI schedules and delivers the micro-coaching dialogue at that time. The example illustrates how an otherwise hidden scheduling decision can become transparent and open to discussion. Alexandre de Montleau will present the PULSE demonstrator in Duisburg.

Two complementary perspectives on adaptive support

Together, the two demonstrators offer a more complete perspective on adaptive digital health interventions: ADHERE asks whether declining engagement calls for additional support; PULSE asks when that support may have the best chance of being received.

Both projects turn adaptive decisions into inspectable and reusable software components. The demonstrations are designed to make these decisions tangible and discussable; they do not yet constitute evaluations of clinical or behavioural effectiveness.

Meet us in Duisburg

Please join Lola Jo Ackermann and Alexandre de Montleau at Mensch und Computer 2026. Visitors will be able to explore synthetic user profiles and interaction histories, vary timing windows and data availability, and see how both systems arrive at their recommendations.

We look forward to discussing adaptive intervention design, engagement, receptivity, adherence, and the responsible use of personalized decision-making with researchers, practitioners, and implementation partners.

The detailed presentation schedule is still being finalized. Please consult the official MuC 2026 programme for the latest information.

References

Piol, G., Ackermann, L. J., Nißen, M., Kowatsch, T. (2026). Designing for Non-Use: Demonstrating Prediction and Prevention of Non-Adherence in Mobile Data Collection and Intervention Studies. Mensch und Computer 2026 – Workshopband, Gesellschaft für Informatik e.V., 30 August–2 September 2026, Duisburg, Germany. https://doi.org/10.18420/muc2026-mci-demo-303. Demo video.

de Montleau, A., Nißen, M., Kowatsch, T. (2026). Demonstrating PULSE with ScrollWise-JITAI: Receptivity-Aware Timing for Just-in-Time Adaptive Interventions. Mensch und Computer 2026 – Workshopband, Gesellschaft für Informatik e.V., 30 August–2 September 2026, Duisburg, Germany. https://doi.org/10.18420/muc2026-mci-demo-304. Demo video.