PDA

View Full Version : Re: Device to measure position/speed during walking andwheelchair use.



Richard Baker
07-09-2006, 12:21 PM
Dear all,



Sorry, its several months since I first posted this request but below is a summary of feedback I had to an enquiry in February about devices to measure position/speed during walking and wheelchair use. Full original posting at foot of e-mail below.



Conclusions:



GPS systems are improving and there are several commercially available units (eg Garmin Forerunner, see link below) which can log speed etc. These are still only reliably outside. Augmented (or assisted) GPS is the next hot thing which uses mobile phone, wireless networks and other RF infrastructure to supply position information when GPS isn’t available. This will still limit monitoring to regions where such infrastructure is available. Even outdoors practical accuracy is of the order of a metre or two and differentiating this to get speed data may introduce a lot of noise.



Systems to specifically track people within a given building are available but generally require a fairly permanent installation for the building concerned.



It is theoretically possible to double integrate acceleration data to obtain position. Such measurements however are subject to considerable drift. Augmenting accelerometer information with magnetometer, barometer and gyroscopic data can reduce such drift. One device (from XSENS see below) claims to have drift free orientation (but not position). Vectronix appear to have developed a similar system to give location for military applications but its been difficult to trace more detailed information.



Most accelerometer based activity monitors analyse the signal from one or more accelerometers to monitor anything from just number of steps, to walking speed and energy expenditure. There are a wide range of devices offering a wide range of specifications. There are a number of devices coming from major manufacturers in the fitness/training/running market based around wristwatch type data loggers with some very impressive looking features at quite reasonable prices. Not clear how these would perform for pathological gait of course!





Links:



GPS Systems:



Garmin Forerunner (http://www.garmin.com/products/forerunner305/) GPS based wrist worn sensor. Logs position, speed etc. Claims to give good accuracy near trees, tall buildings etc.)



Assisted GPS systems. http://www.u-blox.com/





Building monitoring:



Ubisense (http://www.ubisense.net/ ). Uses ultrawideband technology to locate a tag within an indoor environment (approx 30mx30m) with accuracy of about 15cm (velocities may be better due to fixed error in location) using special sensors connected into a building’s existing network. Systems cost around € 8,000.





Accelerometer based orientation and location sensors:



Vectronix (http://www.vectronix.ch ) Seem to be delivering exactly what I’ve specified for military applications but I’ve been unsuccessful in getting any further information out of them or their Australian agents!



XBus/MTx (http://www.xsens.com ). 3DOF orientation tracker based on solid state gyroscopes, accelerometers, magnetometer to give drift free readings. Can be linked to datalogger. Some third party development of sensors and logger to monitor activity but no commercially available software to do this.







Accelerometer based activity loggers:



ActivePAL (http://www.paltechnologies.com/) Small (20g) accelerometer based sensor and data logger most often worn on thigh. Logs time spent standing and walking and number of steps are logged continuously over periods of up to 7 days.



Step Activity Monitor/Stepwatch (http://www.cymatech.com ). Small (38g) accelerometer based sensor and data logger worn on ankle. Logs steps taken each minute for periods of up to two months. Steps can be categorised as low, medium or high intensity.



IDEEA (http://www.minisun.com ) . Small logger (59g) and five even smaller (2g) accelerometer based sensors worn on pelvis, thighs and shanks and connected by wires. Data logging of (32Hz) limited by battery to about 60hours. Software distinguishes over 40 different categories of activity. Can calculate walking speed and an estimate of energy expenditure.



Custom built system from RUSH University Medical Centre (http://www.ortho.rush.edu/tribology/Publications/Abstracts/Rush_Tribology_ASB04_165.pdf) based on inclinometers on thigh and shank and used to identify standing, sitting, walking, stair climbing and lying.



Walkmetre (http://vds-281529.amen-pro.com/pdf/LOCOMETRE.pdf) Lab based system for measuring temporal-spatial characteristics of walking (principles of measurement unclear from sales literature).



Dynaport/Minimod (http://www.mcroberts.nl/start.php?language_id=1 ). 3 DOF accelerometer based signal and datalogger recording at 100Hz for up to 24 hours. Software to analyse signals from pelvis and thigh mounted accelerometers to detect standing, lying, sitting, walking cycling etc.



AMP331 (http://www.dynastream.com/products/amp331/) Accelerometer based data logger worn on ankle.



Triax Elite from Nike. (http://niketown.nike.com/niketown/catalog/pdp.jsp?productId=20510 &categoryId=303436&catalogId=1) Design for running market, logs step count and speed on wrist watch.



Foot pod from Suunto (http://www.suunto.com/suunto/main/product_short.jsp?CONTENT%3C%3Ecnt_id=101341986739 45490 &FOLDER%3C%3Efolder_id=9852723697223649&PRODUCT%3C%3Eprd_id=845524442492074&ASSORTMENT%3C%3East_id=1408474395903593&bmUID=1152497445454) uses wireless technology to communicate running/walking speed from ankle mounted accelerometer to wrist watch type data logger. Developed for fitness industry.



Contributors



Anita Bagley, Sacramento Shriners Hospital.

Chris Batten, Roessingh Research and Development

Charlotte Elsworth, Brookes University

Miguel Velhote Correia, Engineering University of Porto

Netta Harries

Ben Heller, University of Sheffield

Glen Lichtwark, Sarah Beth Williams, Royal Veterinary College, London

Justin Kavanagh, Griffith University

Chris Kirtley

Tony Marsh, Science Wake Forest University

Peter Milburn, University of Otago

J Paul Micallef

William Nechtow,

Dieter Rosenbaum, University Hospital Muenster

Sharon Sonenblum, Georgia Institute of Technology

Ming Sun, MiniSun LLC

Marcel van Hak, Daniel Roetenberg, Xsens Motion Technologies

Susan Wilson, Douglas Maxwell, University of Strathclyde





Original Posting:



Dear all,



We are looking for a device that will log the position of a child or the speed at which they are moving (walking, in a wheelchair or otherwise) over the course of several days. Does anyone have such a device or know of one? We’d be very happy top see this as a collaborative research venture if someone has developed such a device and is seeking an application to try it out on.



BACKGROUND: We’ve got a questionnaire based scale that assesses how well children are able to walk over short (5m), medium (50m) or long distances (500m) and would like to see whether this correlates with how often they actually move over these distances (not necessarily walking, they may be in a wheelchair). I’d assumed we would do this using GPS technology but I now know that this doesn’t work indoors. Obviously we could obtain the information from a device that detects either position (as in the GPS) or speed. (In principal we could integrate an acceleration signal but I’d be concerned about the effect of “drift” in doing this over long periods). This is for a Physiotherapist’s PhD project and I’m really looking for a device we can take “off-the-shelf” with a minimum of development or validation work. We’ve done quite an extensive search of commercially available options and personal contacts and have drawn a blank. We don’t particularly want to use a device that is essentially a step counter firstly because this requires assumptions about the subject’s step length and secondly because it will not pick up wheelchair mobility.



AIM: To record the number of short (3-15m), medium (30-150m) or long trips (300-1500m) made by disabled children over a 4 day period either on foot, in a wheelchair or being carried.



SPECIFICATION:



1. record distance travelled and/or speed (whether on foot, in wheelchair or being carried)
2. log data at 1 second intervals
3. be able to record at least 8 hours at a time without requiring recharging
4. be able to be used over at least 3-4 days
5. operate both indoors and outdoors
6. not too heavy, large or otherwise restrictive.



Data will have to allow us to identify when trips occur and what distance is covered during each one. Software to support this would be extremely useful but could be developed by us.



Richard





Richard Baker PhD CEng CSci

Director Gait CCRE/Gait Analysis Service manager

Murdoch Childrens Research Institute

Royal Childrens Hospital

Parkville 3052, Victoria, Australia



Tel 9345 5354, Fax 9345 5447