AI dashboards could ease the burden of monitoring long-running sensing studies

A prototype helped researchers with daily monitoring and decisions, but also raised questions about how clearly AI systems should explain their suggestions.

Editorial illustration — not from the study.

Long-term studies that collect data from phones or wearable devices need regular checks for technical failures and falling participant compliance. The researchers say existing dashboards provide limited help, leaving monitoring as a time-consuming and demanding task.

The team worked with 13 researchers to identify useful AI capabilities and ways to interact with them, then built the DAIMON prototype. They evaluated it in a task-based study and by deploying it in an ongoing real-world sensing study, using the combined findings to propose design guidelines for future dashboards.

How the dashboard performed

The co-design study identified AI capabilities and interactions that researchers wanted for monitoring longitudinal sensing studies. The DAIMON prototype showed promise in supporting researchers’ day-to-day data monitoring and decision-making tasks in both a task-based evaluation and a deployment within an ongoing real-world study. The studies also surfaced concerns about transparency and about what researchers expect from AI systems. The authors consolidate these findings into design guidelines for AI-augmented dashboards.

Why monitoring support matters

Long-running sensing studies can lose important data when technical problems or declining participant compliance go unnoticed. Dashboards that help researchers monitor these issues and make decisions could make the work more manageable, but the findings indicate that usefulness alone is not enough: researchers also need to understand how AI systems support their outputs and what they should expect from them.

Evidence and open questions

The work combines a co-design study with 13 researchers, a task-based study, and a deployment of the DAIMON prototype in an ongoing real-world sensing study. The abstract does not report the task results in detail, the duration or scale of the deployment, or a quantitative comparison with existing dashboards. It therefore supports claims about perceived promise, workflow design, and identified concerns rather than establishing how much the prototype improves monitoring or study outcomes.

Programmable hardware speeds matrix calculations up to 6.3 times

Researchers tested a programmable hardware design that imitates selected quantum-computing ideas to run matrix calculations in parallel. In representative benchmarks, the design reached up to 6.3 times the speed and 4.1 times the energy efficiency of conventional matrix co-processors and embedded graphics processors.

2 days ago·13 views

AI & Computing

AI and docking agreed in some drug screens, but not all

Researchers evaluated three AI systems on hundreds of molecules bound to a SARS-CoV-2 protein and in three prospective drug screens. The results suggest that combining AI-based predictions with traditional docking may help prioritize candidates, although no single approach worked best in every test.

4 days ago·50 views

AI & Computing

A hardware monitor detected simulated attacks on car networks in 14 microseconds

Researchers developed a hardware-assisted system intended to detect message injection, spoofing, flooding and impersonation attacks on automotive networks. In tests using simulated traffic and a prototype automotive processor system, it reported an average detection time of 14 microseconds and less than 2 percent CPU overhead.

5 days ago·42 views