The Thelma Biotel activity sensor transmitter delivers high-resolution data in the activity range of 0 – 3.465 m / s^2. The activity sensor is fully embedded into the transmitter and uses a 3-axis accelerometer for rapid sampling and monitoring of the changes in acceleration on the transmitter over a chosen sampling duration.


  • Available in all transmitter sizes 6, 7, 9, 13 and 16
  • Long operational lifetime ~ 3 months -> 5 years

The activity sensor is fully embedded in the transmitter with no external ports. It uses a 3-axis accelerometer for rapid sampling to document any activity and movement over a chosen sampling duration.

Thelma Biotel offers the smallest activity sensor acoustic transmitters on the market. It can be delivered down to 6.3 mm in combination with other sensors, such as temperature and depth, with several months of operational lifetime.



The Thelma Biotel activity sensor transmitter is versatile and can be applied in studying animal behaviour in regard to environmental changes, anthropogenic changes, spawning, feeding, activity, migration etc. Depending on the application, the output code can be adjusted for Root mean square (RMS) or Overall dynamic body acceleration (OBDA).


The total acceleration in the activity sensor transmitter is measured in three axes (X, Y and Z) and consists of two components, static and dynamic acceleration. The tag runs the raw acceleration data through a low-pass filter where the static component, with gravity and sensor offset, is subtracted from the acceleration reading. The remaining part of the reading is the dynamic part with information about the animals movement.

The measurement output transmitted can be calculated using the equation for ODBA (overall dynamic body acceleration) or RMS (root mean square). We propose to chose RMS or ODBA based upon the studies your work refernce to, aligning your results for comparison.

“A number of studies have demonstrated a clear individual-specific relationship between oxygen consumption (VO2) and ODBA, showing that ODBA can be used as a proxy for energy expenditure during the tested activities on the tested species” (Wilson et al., 2020). “An important caveat is that acceleration-based proxies only quantify mechanical energy use and cannot account directly for metabolic energy consumed by physiological processes such as thermoregulation” (Gleiss, Wilson, et al., 2011).

Martin Lopez, L. M., Aguilar de Soto, N., Madsen, P. T., & Johnson, M. (2022). Overall dynamic body acceleration measures activity differently on large versus small aquatic animals. Methods in Ecology and Evolution, 13(2), 447-458.


The activity sensor cannot function with standard point measurement, and it must log continuously over a period to catch the activity from the tagged animal. The default setting lets the sensor record acceleration data for 1/3 of the average transmit interval. The measurement starts right after the previous transmit.


From the generic 3-axis acceleration measurements, the tag can for example register the tilt- and roll angle of the fish, as well as motion in the forward- and lateral directions. These key parameters reflects a broad range movement- and activity patterns of aquatic specimens that in turn can be used to study behavior and/or welfare. A key feature of the more advanced use of the activity tag is the capability of detecting one or more motion signatures based on the raw data from the 3-axis accelerometer. The activity tag has a powerful processing unit that can be programmed and adapted to the relevant specimen and movement patterns. Relevant movements include feeding- or spawning behavior, body tilt/orientation, rest- or activity level, comfort- or stress state and much more. The transmitter can for example be programmed to detect attacks against a prey, count how many times this movement signature is detected during a period of time, and transmit the stored number of detected movement signatures wirelessly. These methods is similar to the technology used in your smartwatch or excersise wristband which turns in the screen only when you are turning your arm to look at it.

Such detection algorithms can be tailor made to your specific species and be developed in cooperation with us. The learning should ideally be based on real data from controlled conditions where the relevant specimen is set up to log data directly from the desired motion pattern. In most cases the recording of actual behavior from an animal is not feasible, like spawning or mating behavior which can not be triggered easily. To capture data which can be used to learn and compose a detector for the motion signature, a cabled sensor “tag” is normally used. The cable is plugged directly into the computer and captured data is backed up with video clips. This can be done at your facility directly, or we can do it here at our facility based on video you send us.