Measuring BehaviorOrganisers
- Evrim O. Yilmaz — Donders Institute, Radboud University Medical Center, Nijmegen, The Netherlands
- Judith Homberg — Donders Institute, Radboud University Medical Center, Nijmegen, The Netherlands
- Lucas P.J.J. Noldus — Noldus Information Technology bv and Donders Institute, Radboud University Medical Center, Nijmegen, The Netherlands
Format
This session will start with a symposium of scientific presentations and conclude with an interactive discussion.
Summary
Behavioral and welfare assessment of laboratory rodents in their home cages within the vivarium still relies largely on brief manual observations made at selected times. Such observations provide limited coverage across the day and night and may miss short-lived behavioral or welfare-related events. Automated home-cage monitoring systems make it possible to observe animals continuously in their regular housing and to quantify changes in behavior and welfare over longer periods.
In recent years, several automated home-cage monitoring systems have reached the market. Depending on the application, these systems combine imaging, microphones, telemetry, and environmental sensors. The resulting data can be used to derive measures of locomotor activity, cage use, circadian rhythms, social behavior, ultrasonic vocalizations, body temperature, heart rate, respiration rate, and cage conditions such as temperature, humidity, and ammonia levels. However, individual systems cover only a subset of these capabilities, and each measure must be verified and validated for its intended use. Purpose-built cages can integrate a wider range of sensors and may provide more space for animals to express more natural behavior, whereas retrofit solutions for rack-mounted cages are easier to introduce into current husbandry routines. Together, these technologies can support continuous welfare assessment, earlier detection of welfare concerns, and more timely intervention. Despite these potential benefits, implementation often remains limited to pilot projects or a small number of applications rather than facility-wide use.
A further design choice concerns the required level of monitoring. Some applications require only cage-level monitoring. When individual-level data are needed during group housing, however, each animal must be identified reliably. Current approaches include video-based recognition of applied body markings, camera-readable ear tags with visible codes, and RFID transponders detected by antennas positioned in or around the cage. Signal occlusion, variation in nesting material and close contact between animals can cause identity errors. For welfare monitoring, a key design choice is how rapidly monitoring data can be processed and acted upon. Near-real-time analysis, performed on a local edge device or a remote server, can enable immediate alerts, whereas offline analysis can support retrospective assessment and system validation. The processing architecture also affects latency, storage, bandwidth, maintenance, and heat and noise generation. Wider adoption also requires suitable IT infrastructure, interoperable data formats, transparent analysis methods, manageable data volumes and cost, and compatibility with animal-care workflows. Clear evidence of value and return on investment will also be important. Drawing on experience with these systems, this symposium will assess which measurements and design choices are reliable and feasible in routine use and identify remaining technical and practical limitations. We will bring together technology developers, academic researchers, animal welfare specialists, and industry representatives from pharmaceutical companies and contract research organizations. Participants will present the latest technologies and applications, outline current needs, share lessons from implementation and validation, and define priorities for the next generation of automated home cage monitoring systems.
The speakers will be announced after the review of their short papers has been completed.
