Stair climbing data collection

[Drew Smith]

Dear Readers,

We are planning an observational study that will examine how people ascend stairs. This is not a laboratory-based study; rather we will be observing how ordinary people use stairs 'in the wild' so to speak. We plan to use timing gates to give us some idea as to the velocity that people approach and go up the stairs. We would also like to get some estimates of the clearance of the lead foot and where people place their foot on the first step. Our first thought was to use video, but there would be a lot of data to process just to extract a couple of variables per person. So, our question is: would it be difficult to 'automate' this process: are there programs available that could process the video data (i.e., find the foot and edge of the step) and output the position data we seek? Alternatively, are there any other non-imaging sensors that might be able to give us this information? It's a bit of a long shot (I expect we will have to spend a lot of time taking measurements from video in the end) but perhaps readers have some interesting solutions we could consider.

Thank you!

[Matthew Johnson]

Re: Stair climbing data collection

Dear Drew,

This is a very interesting study. My recommendation is instead of using cameras to measure foot position relative to the edge of the stair, which will be very cumbersome to calculate, I would strongly suggest using force sensor shoe inserts to automate the process. It should simplify your study immensely. The inserts that I would recommend purchasing (I don't know of any companies specifically but I have seen them at many conferences) have a multitude of sensors speckled evenly throughout the insert, so you should be able to tell with a fairly high precision where the foot is relative to the edge of the stair. Plus, you will then have force data in conjunction with all your other metrics.

Regarding clearance...I think you might be stuck with video. You could look into using some type of light gate that has multiple sensors (light beams) at different heights, but I've never heard of one. They have to be out there. There are also biomechanical modeling packages out there (like HM Analyzer from NexGen Ergonomics) that could probably give you the subject's foot rise distance, and then you could simply subtract the height of the stair to obtain clearance.

If you are interested in obtaining metrics like stride length, stride time, range of motion, and other standard gait metrics, email me and I can tell you about my company's products. Or just visit the website in my email signature. Sorry for the self promotion.

I hope this brain dump helps. I really think force shoe inserts are the way to go. I believe they are fairly cheap too.

Regards,

Matthew Johnson
Director of Sales & Marketing
APDM, Inc.
2828 SW Corbett Ave, Suite 130
Portland, Oregon 97201
Direct: 971.506.8166
Office: 503.446.4055
Skype: matt-m-johnson

Visit our website at http://apdm.com

[Drew Smith]

Re: Stair climbing data collection

... I would strongly suggest using force sensor shoe inserts to automate the process. It should simplify your study immensely. The inserts that I would recommend purchasing (I don't know of any companies specifically but I have seen them at many conferences) have a multitude of sensors speckled evenly throughout the insert, so you should be able to tell with a fairly high precision where the foot is relative to the edge of the stair. Plus, you will then have force data in conjunction with all your other metrics.

Regarding clearance...I think you might be stuck with video. You could look into using some type of light gate that has multiple sensors (light beams) at different heights, but I've never heard of one. They have to be out there. There are also biomechanical modeling packages out there (like HM Analyzer from NexGen Ergonomics) that could probably give you the subject's foot rise distance, and then you could simply subtract the height of the stair to obtain clearance.

Mathew, thanks for your suggestions. We will actually be trying to collect data from a public stairway with just ordinary people walking by, i.e., non-subjects. As such, they will not be part of the study in the usual way - recruitment, consent forms, attaching sensors, etc - but since we will only be recording their velocity and foot-step position data, without any identifying information (i.e., we won't record their faces or any other personal data), we think that there are no ethics issues.

I suspect that there may be no way around having someone spend quite a bit of time with quite a lot of video!

[John Burke]

Re: Stair climbing data collection

Hello Drew,

this is not my subject but it sounds very interesting so what about this:

if you can set up an array of beam break devices ( http://www.maplin.co.uk/mk120-ir-bea...detector-27513 ) you could probably also note the angle of travel. These devices are very cheap. The height of the beam on the wall gives you the height of the foot automatically and the travel through the array gives you the angle. An interesting way to get immediate data onto you computer screen might be to remove the Led on each unit and instead connect this across a key on a keyboard. This way each time a beam is broken it registers as a letter. Now all you have to do is search the file for patterns. I have no affiliation with Maplin.

Best wishes

John
as7200@hotmail.com

[Brian Schulz]

Re: Stair climbing data collection

Taking the idea of a foot switch and reversing it a bit- would it be possible to built a false top for each stair with a switch or switches under it? This top could be connected to a battery-powered IR LED that would show up on your video camera, but would be invisible to be the people. This way you could turn the stairs themselves into foot switches by using the IR LED for each stair as the indicator for foot contact, which should enable some level of automated footfall detection using MATLAB image processing toolbox or something similar.

[Ton van den Bogert]

Re: Stair climbing data collection

The spam post has been deleted and user banned. We are seeing more and more spam now and are about to start using the Akismet service with our vBulletin system to block spam. This will hopefully solve the problem.

Whenever spam occurs, please let us know so we can take appropriate action. Easiest is to click on the exclamation mark icon just below the post, to report it. Or just use the "Contact Us" link at the bottom of the page.

Ton van den Bogert
Biomch-L co-moderator

[Drew Smith]

Re: Stair climbing data collection

John and Brian,

Thanks so much for your input. These are terrific ideas and we will explore them. Once we have a plan (and perhaps some pilot data), I will update this thread with our findings.

Ton,

I didn't realise that was the purpose of the exclamation mark icon. I will use that next time. I have deleted my post referring to the spammer.

[Steven Abbott]

Re: Stair climbing data collection

Hello,

Just a thought, but a colleague was using a pressure mat a few years ago for their work and had people walking on it, it showed a perfect imprint of the foot print on a 2d grid, maybe having this on the stairs would work?

Kind regards

Steve Abbott

[Venn Ravichandran]

Re: Stair climbing data collection

If you are just using video, in a public place, would it make it easy to put a grid on the stairs? What I mean is, print out a giant grid, say 0.5inch squares, stick it to the stairs and the wall behind the stairs too. That way the person looking through the video would have an easier time measuring the parameters you are looking for?

[Jeroen Bergmann]

Re: Stair climbing data collection

Dear Drew,

We have investigated stair climbing using a body-worn sensor system. We have published information regarding the leg that people used to start stair climbing. Please see our open access article at http://www.ncbi.nlm.nih.gov/pmc/arti...4/?tool=pubmed. The focus of this particular article was stair ascent. The information regarding stair descent can be found at http://www.xsens.com/en/movement-sci...racking-device. Hope this information might be of use to you.

Kind regards,


Jeroen

--
Jeroen Bergmann, PhD, MSc, BSc, PT
Human Performance Group Researcher
Examination Co-ordinator, Science and Patient (Medicine)
Division of Surgery
Department of Cancer & Surgery
Imperial College London
Room 7L13 Charing Cross Hospital
London W6 8RF
j.bergmann@imperial.ac.uk

[Drew Smith]

Re: Stair climbing data collection

@Steve: we had considered a pressure mat but at the moment, we do not have the budget for this, but thanks for the input!

@Venn: this is also a good idea, but we'd have to also mark the stairs horizontally so we could adjust for parallax error. Thanks for your input!

@Jeroen: we'll file this idea away for further experiments. At the moment, we will just be observing (and measuring) people using the stairs, so we won't be able to attach any sensors or interact with them at all. Thanks for your input!

[Michael Shanok]

Re: Stair climbing data collection

Hello Drew,

Depending upon how complex your study has to be, you might consider taking photographs (or in the alternative capturing individual video frames) and tracing them with a profile from anthropometric software, such as HumanCAD's ManneQuin Pro. The software calculates lots of parameters that would be very tedious and expensive to derive from equipping test subjects with sensors, and you have the advantage of being able to substitute the various body sizes and ethnic proportions offered by the software, as well as change the figure's gender, to study differences. Additionally, you will have anonymous wire frame figures that are easy to cut, paste and print into your final publication.

No matter how you proceed, please let me know when you publish, as stair ascent and descent is a subject that I deal with, constantly.

Best regards,
Mike

[Drew Smith]

Re: Stair climbing data collection

Mike,

Thanks for the input - I think this could be very helpful.

We were planning to conduct lab-based measurements in the future but after consideration of all the issues related to our original plan of making some measurements on naive subjects in situ, we may bring the lab research forward in time and not spend too much time on field research.

I will let you know when we have anything being published.