Physiological and biomechanical assessment of free ranging sports dogs

Lead Research Organisation: Royal Veterinary College
Department Name: Comparative Biomedical Sciences CBS


Recent developments in wireless sensor technology have created new opportunities to make physiological and biomechanical measurements on animals in their natural environment. However, two challenges are faced when deploying wireless sensors on animals. The first is the physical size of the sensor node relative to the animal (eg guidelines for birds are 5% of body mass); the second is handling the potentially large volumes of data generated by sensors such as accelerometers. This size/weight constraint, and the frequent need to deploy over a long timescale means that power consumption (and hence battery life) is important. This can be addressed by using a low power processor and by transmitting the minimum volume of data. To do this one needs to minimise the volume of data transmitted over the radio backbone by: 1. Smart logging, ie monitoring the data from the sensors and discarding it or changing the sampling rate according to the nature of the data. This ranges from not logging movement data when an animal is sleeping to intelligent sleep scheduling, so that the device is powered off for as much of the time as is consistent with obtaining the necessary data. 2. Compression in either the frequency or time domain (latter more likely due to processor power limitations). This is constrained by the requirements of low power and techniques in the time domain such as thresholding and differential encoding. The Toumaz sensor is of particular interest to the applicants because the chip is remarkably low power, and the analogue processing prior to digitisation enables the handling of a diverse range of signals from different sensor types (normally sensors such as strain gauges require additional signal processing circuitry which can be power hungry and physically significant). We seek to develop the hardware and software of the Toumaz sensors for use in two related applications which enable us to collect truly novel biological data. The applications present different challenges for the sensors of battery life, physical size and signal nature; resolving these will allow much wider use of the sensors in future applications. 1. Lameness detection in dogs. Many subtle lamenesses in dogs are only apparent after exercise, but in our experience treadmill exercise of dogs for clinical examination is difficult. We therefore require a system for assessment of lameness during free exercise similar to that we have developed and applied for horses using larger sensors from Xsens (publications by Pfau and Wilson). The Toumaz units appear ideal for this application (due to physical size, battery life, radio protocol and cost) since the real time component and radio synchronisation of data from different units enable real time visualisation of the dog's movement using a computer animation which could be valuable for both diagnosis and teaching purposes (lameness evaluation is an important and challenging aspect of veterinary education and clinical practice). Sensors will record biomechanical and stride data which can be analysed and interpreted to identify lameness. It will also contribute to our knowledge of limitations of running and turning performance. 2. Dynamics of foxhounds - we are interested in how groups of animals interact and physiological drivers of this. We will instrument dogs following a scent trail to record heart rate and stride parameters whilst tracking using GPS. We envisage combining the Toumaz chipset with GPS units developed as part of the EPSRC funded WINES project and our BBSRC CASE award with Forsberg Services. Specific questions relate to the metabolic workload of leading vs following in the pack, which may contribute to understanding of how the pack shares the workload of following the scent. This leads into future applications in the domain of behaviour of packs of animals which is the subject of an EPSRC grant we expect to submit shortly.


10 25 50