Knowing firefighters’ locations in a burning building would dramatically improve their safety. In this study from the Saxion Research Centre for Design and Technology in the Firebee project, an algorithm was developed and tested to enhance the estimation of a person’s location, based on inertial measurements combined with measurements of the earth’s magnetic field. The developed algorithm is an extension of the zero velocity update technique. Without any enhancements, the accuracy of the estimation is in the order of several meters after measuring for only a few seconds. With enhancements, the accuracy improved to be within five meters after measuring for ten minutes. Our result demonstrated that it is possible to determine in which room and on which floor a person is after ten minutes. Major improvements were observed in the estimation of the sensor’s height. The results are promising and the following phases of the project focus on improving the solution and on developing the concept into a practically applicable system.
MULTIFILE
CC-BY-NC-ND This paper was presented at the IADIS Multi Conference on Computer Science and Information Systems MCCSIS2020 There is an increasing interest in indoor occupation and guidance information for business and societal purposes. Scientific literature has paid attention to various ways of detecting occupation using different sensors as data source including various algorithms for estimating occupation rates from this data. Gaining meaningful insights from the data still faces challenges because the potential benefits are not well understood. This study presents a proof-of-concept of an indoor occupation information system, following the design science methodology. We review various types of sensor data that are typically available or easy-to-install in buildings such as offices, classrooms and meeting rooms. This study contributes to current research by incorporating business requirements taken from expert interviews and tackling one of the main barriers for business by designing an affordable system on a common existing infrastructure. We believe that occupation information systems call for further research, in particular also in the context of social distancing because of covid19.
MULTIFILE
Improving estrus detection accuracy could improve sow conception rates,leading to higher production efficiency. Current observation-based estrusdetection practices are labor intensive and less accurate. Around estrus, bodytemperature and activity change. Therefore in this study a telemetric monitoringsystem for body temperature and activity was tested. Firstly Templant2 sensors(TeleMetronics) were validated under lab conditions for temperatures from 35°Cto 45°C, using a water basin with a Julabo heater and a P600 thermometer.Activity measurements were validated with the sensors attached to a stick,simulating sow movements. Secondly, sensors were attached externally to 4gilts and 4 sows for 30 minutes, testing functionality. Thirdly, activity of sowswas recorded manually for 3 days around estrus. Results showed that under labconditions temperature results of sensors, heater and thermometer were highlycorrelated (linear regression, R2=0,96; slope 1,1). Simulated activitiescorresponded consistently with peaks in sensor values. Activity was measuredreliably with the sensor attached externally to the sows. On the farm, sowsshowed more activity (manual observations, P<0.05 for standing up, lying down,sitting down and walking) the day before insemination. We conclude thatmonitoring activity and body temperature is a promising tool for estrousdetection in sows.
LINK
