Introduction of a Measurement System for Quantitative Analysis of

Force and Technique in Competitive Sport Climbing

Dominik Pandurevic, Alexander Sutor and Klaus Hochradel

Institute of Measurement and Sensor Technology, UMIT – Private University for Health Sciences,

Medical Informatics and Technology GmbH, Hall in Tirol, Austria

Keywords:

Sport Climbing, Sport Science, Digitalization in Sports, Force Sensors, Markerless-Motion-Capturing.

Abstract:

Rapid development and progress in competitive sport climbing lead to increasing media attention and appre-

ciation in the sports world. Therefore, the increase of requirements to climbing athletes are unavoidably. The

next step to replace existing and objective training methods, is the digitalization in terms of development of

suitable measurement systems. By determining the distribution of forces as well as analysing technique, the

presented method enables the evaluation of athlete’s performance. With the aid of this combined system -

composed of multiple 3-axis high-precision force sensor embedded climbing holds and a marker-less motion

capturing framework - and a depending on the needs and application designed user-interface, coaches get the

opportunity to evaluate their athlete’s training outcomes. This sensor technology allows us to detect mag-

nitude and direction of measured force vectors and the associated distribution of the body weight on hands

versus feet. Additionally, the recording of athlete’s motion including depth information enables calculation of

relevant joint angles and body’s centre of gravity. This measurement system, its modiﬁcation and base com-

ponents, respectively, could be used among others for dynamic move analysis in bouldering or video analysis

in speed climbing.

1 INTRODUCTION

The progress of sport climbing now and in the last

25 years leads to necessary changes to meet the re-

quirements set to athletes, especially for the disci-

plines bouldering and speed climbing. Hence, train-

ing methods in a save and deﬁned environment are

ensured by the spread of artiﬁcial climbing gyms

all over the world, which also allow highly special-

ized and individual training units. By the participa-

tion at the Olympic games in Tokyo 2021, climbing

reaches its historical peak as competitive sport. The

progress of competition by introducing a combined

mode consisting of the three disciplines bouldering,

speed and lead climbing demands versatile athletes

and the change of existing training methods from the

bottom up. Nevertheless, the evaluation of athlete’s

trainings mainly consist of video analysis and the

personal experience of coaches. To reach the state

of the art of other mass sports like cycling or run-

ning, a novel method is herewith presented to help

coaches and athletes to evaluate and monitor their per-

formance.

The typical physique of a climbing athlete differ-

entiates massively from other sportsmen. Therefore

several researches about the performance and asso-

ciated factors of the slim and light constitution of

climbers were done by Giles et al. (2006). Besides ex-

isting methods like the determination of the maximum

ﬁnger force and the correlation with the degree of dif-

ﬁculty (Ferguson and Brown, 1997; Grant et al., 2001,

1996; Mermier et al., 2000), there are no approved

systems for the measurement of magnitude and direc-

tion of absolute forces.

Because of the rapid development of sport climb-

ing and the increasing attention, the digitalization of

trainings methods and the enhancement of the perfor-

mance is unavoidable and therefore topics of various

research projects. A Swiss research group worked

on instrumented climbing holds for the realization

of bouldering routes and the analysis of repeatabil-

ity of climbing patterns (Donath and Wolf, 2015).

For that reason they used wired 6-axis force sensors,

which were mounted on cover disks and inserted into

a climbing wall by cutting out holes with appropriate

diameter. As a result they calculated parameters like

contact time or force maximum to compare them with

Pandurevic, D., Sutor, A. and Hochradel, K.

Introduction of a Measurement System for Quantitative Analysis of Force and Technique in Competitive Sport Climbing.

DOI: 10.5220/0010010001730177

In Proceedings of the 8th International Conference on Sport Sciences Research and Technology Support (icSPORTS 2020), pages 173-177

ISBN: 978-989-758-481-7

173

different trials and athletes.

Another important aspect of analysis in climbing

sport is the recording of climbing motions to calcu-

late relevant parameters such as joint angles or cen-

tre of gravity, which will replace the manual anal-

ysis of video footage. Researches of the university

of Pisa analysed kinematics of a technique used by

most of all rock climbers (Artoni et al., 2017). They

constructed an experimental setup in the laboratory to

simulate the so-called lolotte technique. Therefore,

they built a steel scaffolding with correctly placed

common climbing holds and set up multiple infrared

cameras to capture the motion using marker equipped

athlete. The measured coordinate data of the joints

where used as input for a biomechanics analysis soft-

ware to determine the effects on knee and hip caused

by this technique. This highly specialized research

project estimates the impact on joints by using the

lolotte technique and is generally not deployable in

climbing.

A method described in (Iguma et al., 2020) com-

bines ﬁve 6-axis force sensors with seven 3D motion

capture cameras to a measurement system for analy-

sis of human motions and the corresponding forces.

Compared to the here described system, the Japanese

researcher group works with multiple reﬂective mark-

ers on the body for high precise tracking of climb-

ing motions. Hence, they were able to determined the

motion conditional forces by calculating the resulting

acceleration of the body’s centre of gravity and com-

pared them to the force sensor’s output. Additionally,

they analysed and compared the centre of mass tra-

jectories of beginners and experienced climbers. Be-

cause of the elaborate setup of this measurement sys-

tem, the usage beyond research purposes is barely re-

alizable.

Because of the variety of athletes and the different

disciplines, the ﬂexibility of this measurement system

is essential. Therefore, a collaboration with the Aus-

trian Climbing Federation and its trainers and athletes

will make a development of training methods with

specialized requirements possible. This paper will

present a method for describing a measurement sys-

tem used for bouldering training.

2 METHOD

Compared to speed climbing, an entire evaluation of

the performance of bouldering athletes is very difﬁ-

cult due to the variety of techniques and complexity

of movements. In spite of everything there are still

important and measurable parameters, which can be

determined by existing, nowadays default measure-

ment equipment. On the other hand, there are also

magnitudes like athletes conditional parameters (ape

index), outside inﬂuences (temperature) or unmea-

surable, psychological impacts, that should be deter-

mined to complete the hereafter introduced system.

Anyway, this paper will only focus on the measure-

ment system consisting of force and motion determi-

nation. Therefore, the presented method will be split

up in these two ﬁelds, where each one will be de-

scribed from its setup to the usage.

2.1 Force Measurement

To implement the idea of the distribution of forces on

hands and feet, a wireless force measurement system

for easy integration onto artiﬁcial climbing walls is

presented.

Figure 1: K3D120 3-axis force sensor with mounting box

and hold.

As seen in Figure 1, the core of each instrumented

climbing hold is the high precise 3-axis force sen-

sors K3D120 from ME-Meßsysteme, which are used

to measure the 3D force vector in a range of ±1kN

for any axis. To protect it from outer impacts each

one is surrounded by pieces of wood. At the back it

is ﬁxed on a mounting plate for an simple integration

on climbing walls.

As mentioned before, the structuring of a com-

plete wireless measurement system is essential to en-

able a simple installation on the walls and uncom-

plicated execution of trials. Therefore, the climb-

ing holds are equipped with battery powered micro-

controllers including a WiFi module. To ensure syn-

chronicity and communication between each other, it

beneﬁts from the use of the MQTT broker/client prin-

ciple. Hence, it is able to receive commands from a

deﬁned communication interface and send measured

forces and other current states.

2.2 Motion Measurement

The combination of the force distribution on hands

and feet and the motion of the climber expressed by

icSPORTS 2020 - 8th International Conference on Sport Sciences Research and Technology Support

174

the trend of joint angles as well as the position of the

center of gravity and its distance to the wall ensures

holistic information for analysing climbing perfor-

mances. Hence, with the aid of the Realsense D435

depth camera from Intel, this system is used for the

integration of the so-called Marker-less Motion Cap-

ture Method. As the title implies, the measurement

and calculation of motion-speciﬁed parameters take

place without any use of markers on the body of the

athlete. Therefore, the athlete is not affected or dis-

turbed by this system.

To implement the idea of marker-less motion cap-

turing the depth camera is used to receive RGB and

depth frames. After aligning them the resulting data

is processed by the real-time multi-person software

OpenPose (Cao et al., 2018) for the detection of hu-

man body keypoints on images or video ﬁles. By a

suitable arrangement of one or more cameras it is pos-

sible to analyse the recorded motion pattern (see Fig-

ure 2).

Figure 2: Example for Marker-less Motion Tracking.

Also post-processing of already recorded training

trials or competitions is possible with this method.

This is especially very useful for the automated anal-

ysis in speed climbing, which replaces processing by

coaches. Additionally, the straightforward setup con-

sisting of only one depth camera and the omission of

any calibration and use of markers on the body en-

ables non-invasive measurements, which incredibly

increases the quality of training evaluations.

Because of the vast amount of data receiving from

one trial, the focus for this paper is on the determi-

nation of the body’s center of gravity and its distance

to the wall as well as the distribution of the athlete’s

weight on hands and feet. For the calculation of the

centre of gravity the weights of the single body parts

and the location of their mass centres are needed.

With the parameters of Table 1 the body’s centre C

for every frame can be determined with

C =

∑

i=1

· x

, (1)

where x

is calculated with the coordinates of the by

OpenPose (Cao et al., 2018) identiﬁed joints and the

corresponding radii r

as follows:

= s

· r

+ j

p,i

. (2)

The variables s

and j

p,i

are deﬁned as the vector and

the proximal joint of the i-th body segment.

With the recorded depth information, the hence

ﬁtted plane of the climbing wall and its normal vector

are used to calculate the distance of the body centre

to the wall with the dot product

d = (C

C − P

P) · n

n, (3)

with P

P as any point on the plane and n

n as the normal

vector.

Table 1: Distribution of weight and mass center radius for

each body segment (de Leva, 1996).

body segment mass m

[%] radius r

[%]

Head 6.94 50.02

Trunk 43.46 44.86

Upper Arm 2.71 57.72

Forearm 2.23 67.51

Upper Leg 14.16 40.95

Lower Leg 4.33 44.59

Foot 1.37 44.15

For the results presented in the following section,

both subsystems measured with a sample rate of 30

Hz. Since both measurement units are recording in-

dependently, the outcomes need to be synchronized.

By logging timestamps of the operating system for

the force sensors and the depth camera and assuming

they run with a accuracy of ± 0.5 Hz, it sufﬁces to

compare the timestamps at the starting point of each

recording and shift the delayed data set by the result-

ing time lag.

3 RESULTS

The above-mentioned measurement system consist-

ing of eight climbing holds with integrated 3-axis

force sensors and a marker-less motion capturing sys-

tem including a depth camera, was ﬁrst applied in the

climbing gym ’Kletterzentrum Innsbruck’ for valida-

tion purpose.

Therefore, the holds were placed on a boulder-

ing wall at speciﬁed positions to analyse two differ-

ent motion patterns (see Figure 3). The depth cam-

era was positioned such that the starting pose as well

as the whole climbing route would be visible on the

recordings. The orientation of the camera was cal-

culated using the built-in accelerometer for determi-

nation of the disposition to the wall. Three athletes

with related potential were chosen to realise initially

comparisons between them with the aid of body pa-

rameters (height, weight, ape-index).

In Figure 3 you can also see, that by synchroniz-

ing data of force measurement and motion capturing

Introduction of a Measurement System for Quantitative Analysis of Force and Technique in Competitive Sport Climbing

175

Figure 3: Snapshot of a climbing motion with illustration of

force of one hold and the joint angle of both knees.

Figure 4: Distribution of the distance of the body centre to

the wall for a selected area of samples.

systems, the resulting output contains a plenty of in-

formation about one climbing trial. A great beneﬁt

of this system is the complete wireless and marker-

less solution. Therefore the athlete is free of thoughts

about any measurements. An example for the result-

ing characteristics of the distance to the wall for a se-

lected range of samples is apparent in Figure 4.

The other important parameter to superﬁcial rea-

son the performance of athletes is the ratio of the force

distribution on hands and feet. Compared to the cal-

culation of the athlete’s distance to the wall, where

only the motion tracking system is used, both parts

of the above presented measurement system are as-

signed. Therefore the integrated climbing holds are

used to get the force vector for every sample. Since

the recording of the force distribution is done for the

Figure 5: Relationship between body centre’s distance to

the wall and the ratio of force on hands and feet.

whole climbing trial, the extraction of portions, where

the four body segments were grabbing/booting the

corresponding hold, is required. At this point the

recordings of the depth camera are employed.

First, the location of the position of the embedded

climbings holds on the wall is accomplished. Con-

sidering that a static snapshot without any athlete was

gathered to mark the holds and a boundary rectangle

with deﬁned threshold around them, respectively. By

using the detected joints, a implemented algorithm is

able to recognize the moment of entering one of these

threshold boxes by a body segment of interest and

mapping them. The resulting characteristics of hands

and feet, respectively, are vectorial added and normal-

ized for every sample. The quotient of them returns

the hands-versus-feet-ratio

R =

+ F

with F

, F

as the force vectors of

hands and feet. An indicator for good and energy efﬁ-

cient performance is a ratio R < 1, whereby the body

weight is shifted towards the feet.

To ensure this condition a decrease of the distance

of the body’s centre of gravity to the wall is required.

In Figure 4 and 5 a correlation of this value and the

ratio R is visible. Figure 5 represent the relationship

between them and marks the efﬁcient samples with

green and the poorer ones with red dots. Except of

few outliers the condition R < 1 is met within a re-

duction of the body centre’s distance to the wall.

icSPORTS 2020 - 8th International Conference on Sport Sciences Research and Technology Support

176

The calculation of the distance of the athletes cen-

tre of gravity to the wall and the associated measure-

ment results are highly dependent of accuracy of the

depth information of the used camera system. There-

fore for these ﬁrst climbing trials, the optimal posi-

tioning of the depth camera was essential to avoid en-

vironmental inﬂuences like light and reﬂections.

4 CONCLUSION

Herewith a new method to open up new perspectives

for the speciﬁc force and technique evaluation in com-

petitive sport climbing was presented. This system

implements simultaneous measurements of force and

motion with the possibility to evaluate athlete’s per-

formance by analysing parameters like the distance

of mass centre of the body to the wall or the hands-

versus-feet-ratio. On the other hand you can also

use the measured force vectors and captured joint an-

gles for the study of early identiﬁcation of injuries

or fatigue by evaluating the repeatability of multiple

climbing trials. Minor modiﬁcations of the existing

system (enlargement of the sensors housing for the

use of bigger holds, usage of a mobile tracking cam-

era system, etc.) enable an enhancement of the area of

application. Of course, systems using a marker-based

measuring system increase the accuracy of motion-

dependent parameters. The non-use of markers on

the athlete’s body prevents invasion on their technique

while measurements, but also creates a dependency

of the used camera system and its positioning. On

that account the above mentioned modiﬁcations will

be implemented as well as the usage of the lidar (light

detection and ranging) camera Realsense L515 from

Intel to get among other things a better accuracy of the

depth information, which is essential for the validity

of this measurement system.

Working in a cooperation with the ’Kletterzen-

trum Innsbruck’ and the Austrian Climbing Federa-

tion and its national trainers on two further projects

in the ﬁeld of bouldering and speed climbing will

also lead to a advancement of the described method.

Therefore the presented measurement system is modi-

ﬁed in such way, that e.g. equipping larger holds with

multiple connected force sensors could enabled dy-

namic move analysis (’Dyno’, jumps).

REFERENCES

Artoni, A., Tomasi, M., and Di Puccio, F., editors (2017).

Kinematic Analysis of the Lolotte Technique in Rock

Climbing, volume Volume 6: 13th International Confer-

ence on Multibody Systems, Nonlinear Dynamics, and

Control of International Design Engineering Technical

Conferences and Computers and Information in Engi-

neering Conference.

Cao, Z., Hidalgo, G., Simon, T., Wei, S.-E., and Sheikh,

Y. (2018). OpenPose: realtime multi-person 2D pose

estimation using Part Afﬁnity Fields. In arXiv preprint

arXiv:1812.08008.

de Leva, P. (1996). Adjustments to zatsiorsky-seluyanov’s

segment inertia parameters. Journal of Biomechanics,

29(9):1223–1230.

Donath, L. and Wolf, P. (2015). Reliability of force appli-

cation to instrumented climbing holds in elite climbers.

Journal of applied biomechanics, 31(5):377–382.

Ferguson, R. A. and Brown, M. D. (1997). Arterial blood

pressure and forearm vascular conductance responses to

sustained and rhythmic isometric exercise and arterial

occlusion in trained rock climbers and untrained seden-

tary subjects. Eur. J. Appl. Physiol. Occup. Physiol.,

76(2):174–180.

Giles, L. V., Rhodes, E. C., and Taunton, J. E. (2006). The

physiology of rock climbing. Sports Med. (Auckland,

N.Z.), 36(6):529–545.

Grant, S., Hasler, T., Davies, C., Aitchison, T. C., Wilson,

J., and Whittaker, A. (2001). A comparison of the anthro-

pometric, strength, endurance and ﬂexibility characteris-

tics of female elite and recreational climbers and non-

climbers. J. Sports Sci., 19(7):499–505.

Grant, S., Hynes, V., Whittaker, A., and Aitchison, T.

(1996). Anthropometric, strength, endurance and ﬂex-

ibility characteristics of elite and recreational climbers.

J. Sports Sci., 14(4):301–309.

Iguma, H., Kawamura, A., and Kurazume, R. (12.01.2020

- 15.01.2020). A new 3d motion and force measurement

system for sport climbing. In 2020 IEEE/SICE Inter-

national Symposium on System Integration (SII), pages

1002–1007. IEEE.

Mermier, C. M., Janot, J. M., Parker, D. L., and Swan,

J. G. (2000). Physiological and anthropometric determi-

nants of sport climbing performance. Br. J. Sports. Med.,

34(5):359–65; discussion 366.

Introduction of a Measurement System for Quantitative Analysis of Force and Technique in Competitive Sport Climbing

177