This is a sel -a chi ed e sion o an o iginal a icle. This e sion
may di e om he o iginal in pagina ion and ypog aphic de ails.
Au ho (s):
Ti le:
Yea :
Ve sion:
Copy igh :
Righ s:
Righ s u l:
Please ci e he o iginal e sion:
CC BY 4.0
h ps://c ea i ecommons.o g/licenses/by/4.0/
The E ec o Paddle S oke Va iables Measu ed by T ainesense Sma Paddle® on he
Veloci y o he Kayak
© 2022 by he au ho s. Licensee MDPI, Basel, Swi ze land.
Published e sion
Löppönen, An i; Vän inen, Tomi; Ha e inen, Ma ko; Linnamo, Vesa
Löppönen, A., Vän inen, T., Ha e inen, M., & Linnamo, V. (2022). The E ec o Paddle S oke
Va iables Measu ed by T ainesense Sma Paddle® on he Veloci y o he Kayak. Senso s, 22(3),
A icle 938. h ps://doi.o g/10.3390/s22030938
2022


Ci a ion: Löppönen, A.; Vän inen, T.;
Ha e inen, M.; Linnamo, V. The
E ec o Paddle S oke Va iables
Measu ed by T ainesense
Sma Paddle®on he Veloci y o he
Kayak. Senso s 2022,22, 938.
h ps://doi.o g/10.3390/s22030938
Academic Edi o : Giuseppe Vannozzi
Recei ed: 10 No embe 2021
Accep ed: 23 Janua y 2022
Published: 26 Janua y 2022
Publishe ’s No e: MDPI s ays neu al
wi h ega d o ju isdic ional claims in
published maps and ins i u ional a il-
ia ions.
Copy igh : © 2022 by he au ho s.
Licensee MDPI, Basel, Swi ze land.
This a icle is an open access a icle
dis ibu ed unde he e ms and
condi ions o he C ea i e Commons
A ibu ion (CC BY) license (h ps://
c ea i ecommons.o g/licenses/by/
4.0/).
senso s
A icle
The E ec o Paddle S oke Va iables Measu ed by T ainesense
Sma Paddle®on he Veloci y o he Kayak
An i Löppönen 1,2 , Tomi Vän inen 3, Ma ko Ha e inen 1,4 and Vesa Linnamo 1,*
1Facul y o Spo and Heal h Sciences, Uni e si y o Jy äskylä, 40014 Jy äskylä, Finland;
[email p o ec ed] (A.L.); [email p o ec ed] (M.H.)
2Physical Ac i i y, Spo s and Heal h Resea ch G oup, Depa men o Mo emen Sciences, KU Leu en,
3000 Leu en, Belgium
3KIHU-Resea ch Ins i u e o Olympic Spo s, 40700 Jy äskylä, Finland; [email p o ec ed]
4Va ala Spo s Ins i u e, 33240 Tampe e, Finland
*Co espondence: [email p o ec ed]; Tel.: +358-405044800
Abs ac :
(1) Backg ound: This s udy aimed o compa e key a iables o paddle s oke measu ed
by a comme cial T ainesense Sma Paddle
®
agains he s ain-gauge sha and in es iga e how
hese a iables a e associa ed wi h he eloci y o he boa among na ional-le el canoe polo playe s.
(2) Me hods: This
s udy in ol ed 14 Finnish na ional-le el canoe polo playe s. The measu emen
p o ocol consis ed o h ee di e en paddling eloci ies, which we e pe o med in indoo swimming
pools. The eloci y o he boa was calcula ed based on he pe o mance ime measu ed wi h he
lase pho ocell ga e. Canoe polo equipmen was used in he s udy and a Sma Paddle senso was
a ached o he paddle blade. A s ain-gauge paddle sha was used as a e e ence me hod o examine
he alidi y o Sma Paddle. (3) Resul s: The s oke a e, o ce p oduc ion ime, mean and maximal
o ce measu ed wi h he s ain-gauge paddle sha co ela ed s ongly ( = 0.84–0.95, p< 0.01) wi h
Sma Paddle. Howe e , he Sma Paddle o e es ima ed he maximum o ce compa ed o he s ain-
gauge sha . S oke a e ( = 0.86, p< 0.01), mean o ce ( = 0.79, p< 0.01), maximal o ce ( = 0.78,
p< 0.01) and o al absolu e impulse ( = 0.70, p< 0.01) co ela ed posi i ely and o ce p oduc ion
ime nega i ely ( =
−
0.76, p< 0.01) wi h he eloci y o he boa . (4) Conclusions: We conclude
ha he Sma Paddle p o ides p omising in o ma ion on s oke key a iables when compa ed o
he s ain-gauge paddle sha . The Sma Paddle is a new and in e es ing ool o biomechanical
esea ch and daily kayaking coaching in eal open wa e condi ions. Howe e , mo e esea ch and
algo i hm de elopmen a e needed be o e he Sma Paddle can be used in e e yday coaching sessions
in kayaking.
Keywo ds: kayaking; biomechanics; canoe polo; ine ial measu emen uni
1. In oduc ion
To c ea e a o wa d accele a ion o he boa , he paddle mus win he esis ance
o ces o wa e and ai [
1
–
4
]. Fo ce, powe , and s oke a e a e he mos impo an ac o s
in luencing he eloci y o he boa [
5
–
9
]. In addi ion, i is impo an o p oduce a “bell-
shaped” o ce p o ile, which leads o a bigge impulse [
6
,
10
,
11
]. The paddle s oke is mos
e ec i e when he blade is pe pendicula o he wa e su ace [
2
,
12
] and he maximum o ce
is eached jus be o e his [
13
,
14
]. S udies ocused on he paddle ajec o y ha e shown ha
mo e expe ienced paddle s a e capable o mo e symme ical paddling pe o mance [
15
–
17
]
and ha e wide ajec o ies han inexpe ienced paddle s [16].
T adi ionally, he di ec ion o he paddle s oke is pa allel o he kayak’s cen e line,
bu on a wing paddle, he ajec o y is sligh ly di e en because he blade’s wing shape
p oduces buoyancy compa ed o a adi ional paddle [
11
,
12
,
15
]. When he paddle is using
he wing paddle, he ajec o y should be on a sligh ly mo e cu ed line, mo ing beyond
Senso s 2022,22, 938. h ps://doi.o g/10.3390/s22030938 h ps://www.mdpi.com/jou nal/senso s
Senso s 2022,22, 938 2 o 12
he cen e line o he boa [
15
]. This has been ound o p olong he op imal du a ion o he
p opulsion gene a ion posi ion [
11
]. The s oke ajec o y is hus one o he key ac o s
ela ed o paddling pe o mance and a ies be ween di e en ypes o paddling such as
canoe polo and sp in kayaking. Al hough he canoe polo is paddled wi h a adi ional
paddle ( la -blade) and should ollow he same p inciples as o he kayak spo s using a
adi ional paddle, he ac ual esea ch e idence on canoe polo is s ill limi ed.
Mos o he biomechanical kayaking s udies ha e been conduc ed h ough mo ion
analysis and di ec o ce measu emen s. S udies ha e been made by e gome e s in labo a-
o y condi ions [
8
,
18
–
20
] and open wa e en i onmen s [
7
,
10
,
12
,
21
–
23
]. S udies in an open
wa e en i onmen o en use a s ain-gauge paddle sha [
6
,
21
,
23
] o measu e he bending
o he sha and he eby he a iables o he s oke. This me hod has also been ound o
be a eliable analyze on-wa e owing pe o mance [
24
]. Paddle blade ajec o y- ela ed
s udies ha e p e iously been done wi h an e gome e in labo a o y condi ions [
20
,
25
] and
open wa e using cinema og aphic came as [12,26].
Howe e , he limi a ion o hese me hods is ha hey canno be used o s udy he di ec-
ion o he paddle s oke o he ajec o y o he paddle blade when paddling in open wa e
condi ions o long dis ances (compe i i e analyzes). Mo eo e , many me hods such as 3D-
mo ion analysis a e no sui able o di ec ans e o an open wa e en i onmen especially
o daily ( echnical) coaching sessions, whe e a s and-alone sys em is o en equi ed.
The new compac senso echnology could o e new oppo uni ies o biomechanical
esea ch and daily coaching in wa e spo s. The small-sized ine ial measu emen uni s
allow hese de ices o be used in eal condi ions [
27
–
29
]. Ine ial measu emen uni s a e
capable o measu ing, p ocessing, and s o ing da a in a compac size and a e also e y ligh
and a e hus possible o in eg a e in o equipmen [
27
]. Ine ial measu emen uni s ha e
been used in wa e spo s especially in swimming [
30
,
31
] bu also in kayaking o measu e
kayak mo emen s [
32
,
33
]. I is also no ewo hy ha he de elopmen o algo i hms plays
a key ole as senso echnology is a g owing me hod in he biomechanical measu emen
o spo s. Fo example in swimming, s oke a e, swimming eloci y, and s oke phases
ha e been he mos impo an a iables in algo i hm de elopmen [
34
]. Howe e , o ou
knowledge, he ine ial measu emen uni s ha e no been used o measu e simila kayaking
s oke a iables such as o ce, s oke, o o ce p oduc ion ime. The ine ia measu emen
uni could p o ide a ool o es ima e he di ec ion o paddle s oke unde eal condi ions
because he da a measu ed by i s senso s can be used o de e mine ajec o ies.
In his s udy, we use he T ainesense Sma Paddle
®
which is an ad anced comme cial
wea able unde wa e senso . I has been widely used in swimming o examine echnique
( ajec o y, eloci y, and o ien a ion o he hand) and he applied o ces du ing a swimming
s oke [
35
]. This s udy aimed o compa e key a iables o paddle s oke measu ed by
a Sma Paddle agains he s ain-gauge paddle sha . In addi ion, we in es iga e how
hese key a iables a e associa ed wi h he eloci y o he boa among na ional-le el canoe
polo playe s.
2. Ma e ials and Me hods
2.1. S udy Design and Pa icipan s
This c oss-sec ional s udy consis ed o wo da ase s. The i s da ase (n= 6, all
we e men) compa ed he esul s o he Sma Paddle and a s ain-gauge paddle sha in
key paddle s oke a iables. The second s udy ( o al n= 14, men n= 12, women n= 2)
in es iga ed he e ec o key paddle s oke a iables measu ed on a Sma Paddle on he
eloci y o a boa .
The pa icipan s o his s udy we e Finnish na ional-le el canoe polo playe s (Table 1).
The a e age age o he da ase one pa icipan s (n= 6) was 39.3
±
10.1 yea s, heigh
183.3 ±10.6 cm
, and weigh 81.3
±
13.0 kg. The a e age angle o he blade was 80.0
±
0.0
◦
and he paddle leng h was 206.0
±
0.0 cm. The a e age age o he da ase wo pa icipan s
(n= 14) was 40.8
±
17.2 yea s, heigh 182.3
±
7.6 cm, and weigh 81.2
±
11.6 kg. The
a e age angle o he blade was 69.1 ±17.9◦and he paddle leng h was 206.4 ±0.9 cm.
Senso s 2022,22, 938 3 o 12
Table 1. The desc ip i e in o ma ion o he s udy pa icipan s.
Mean (SD) 1s Da ase (n= 6) 12nd Da ase (n= 14) 1
Age [yea s] 39.3 ±10.1 40.8 ±17.2
Heigh [cm] 183.3 ±10.6 182.3 ±7.6
Weigh [kg] 81.3 ±12.9 81.2 ±11.6
Blade angle [deg] 80.0 ±0.0 269.1 ±17.9
Paddle leng h [cm] 206.0 ±0.0 2206.4 ±0.9
SD = S anda d De ia ion,
1
Desc ip i e able is missing da a om i e pa icipan s because hey did no eply o
he ques ionnai e sen a e wa d (1s da ase = 2 pa icipan s, 2nd da ase = 3 pa icipan s). 2no adjus able.
The inclusion c i e ion o he pa icipan s we e: (1) a leas h ee yea s o expe ience
o compe i ion kayaking, (2) basic heal h, and (3) we e no allowed o ha e any inju ies
o o he p oblems ha in e e ed wi h hei kayaking pe o mance. The s udy p o ocol
ollowed he p inciples o he Decla a ion o Helsinki. The s udy was app o ed by he
Human Sciences E hics Commi ee o he Uni e si y o Jy äskylä. All pa icipan s signed
w i en in o med consen be o e pa icipa ing in he s udy.
2.2. Measu emen
Measu emen s we e pe o med in h ee di e en indoo swimming pools, in wo
loca ions in Finland. All pa icipan s in da ase one we e measu ed in a 25-m pool wi h a
measu emen dis ance o 15 m. Fou pa icipan s o da ase wo we e measu ed in a
50-m
pool (measu ing dis ance 30 m) and he emaining 10 pa icipan s we e measu ed in a 25-m
pool (measu ing dis ance 15 m). Based on he paddling eloci ies, he measu emen dis ance
did no a ec he esul s o he analysis. A a 30-m analysis dis ance (50-m pool), mean (SD)
13.9
±
1.7 s okes pe side we e included in he analysis a eloci y one,
14.6 ±2.4 s okes
a eloci y wo, and 15.5
±
2.1 s okes a eloci y h ee. A a 15-m analysis dis ance (25-m
pool), 6.4
±
0.8 s okes we e included in he analysis a eloci y one, 6.9
±
1.0 s okes a
eloci y wo, and 7.1
±
0.9 s okes a eloci y h ee. The eloci y o he canoe was calcula ed
based on he pe o mance ime measu ed wi h he lase pho ocell ga es placed on he edges
o he pool. The dis ance be ween he pho ocells was measu ed and he a e age eloci y
was calcula ed by di iding he paddled dis ance by he pe o mance ime ( = s/ , whe e
= eloci y
, s = dis ance, and = ime). The pho ocells we e placed in he pools so ha he
boa could be accele a ed (accele a ion dis ance was 5 m) o he c uising eloci y be o e he
i s pho ocell ga e s a ed he iming.
The measu emen p o ocol o da ase one (Sma Paddle & s ain-gauge sha compa -
ison) consis ed o h ee di e en eloci ies o paddling pe o mances. Fi s pa icipan s
we e ins uc ed o paddle a a basic endu ance pace ( eloci y 1), which was desc ibed as a
slow and calm bu s ong pace ha can be main ained o a long ime. The second eloci y
was clea ly ( eloci y 2) as e paddling, which was desc ibed as s ong, bu echnically
pu e, paddling. This eloci y was pe o med wi h h ee di e en echniques (no mal,
ene ge ic bea ing s oke, delayed long and ligh s oke) o add he e ogenei y o he da a
used in he alidi y analysis. The las was maximum eloci y ( eloci y 3), which was
desc ibed as a nea maximum pe o mance, whe e he pace is almos a maximum, bu he
paddling echnique emains high quali y. In o al, da ase one con ained a o al o 60 da a
poin s (=6 pa icipan s
×
5 di e en pe o mances/pa icipan
×
2 sides (le and igh ).
The eloci ies
one (1.95
±
0.22 m/s) wo (2.20
±
0.21 m/s) and h ee (2.50
±
0.20 m/s) di -
e ed signi ican ly (p< 0.05) as planned. The eloci ies o he wo di e en echnical s yles
(ene ge ic bea ing s oke 2.36
±
0.26 m/s & delayed long and ligh s oke
2.25 ±0.21 m/s
)
we e he same as no mal s yle o eloci y wo (p= 0.313, p= 0.563. espec i ely) as planned.
Da ase wo consis ed o he h ee di e en eloci ies wi h no mal paddling pe o -
mances men ioned abo e ( eloci y 1, 2 & 3), o s udy he e ec o key paddle s oke a iables
measu ed on a Sma Paddle on he eloci y o a boa . Da ase wo con ained a o al o
84 da a poin s
(=14 pa icipan s ×3 di e en eloci y pe o mances/pa icipan ×2 sides
Senso s 2022,22, 938 4 o 12
(le and igh )). The eloci y o he boa di e ed signi ican ly be ween di e en eloci ies
(p< 0.001) as planned.
2.3. Ins umen a ion
The comme cial Sma Paddle (T ainesense Oy, Tampe e, Finland) was used as a
measu ing uni (9-axis IMU + p essu e senso ). The de ice uses a sampling equency o
100 Hz and he da a was s o ed in he T ainesense Analysis Cen e cloud se ice. The da a
was uploaded o he mobile phone and onwa d Analysis Cen e ia a Blue oo h connec ion.
The Sma Paddle was moun ed on he back o he blade (B aca-Spo Polo Kine ic max,
adjus able sha , ze o-leng h 205 cm) using silicone s aps, and a 15 mm diame e hole was
d illed in he paddle blade o he p essu e senso (Figu e 1).
Senso s 2022, 22, x FOR PEER REVIEW 5 o 13
Figu e 1. (A) Sma Paddle place on he blade, (B) S ain-gauge sha in calib a ion boa d (wi hou
IMU a ached), (C) Sma Paddle a ached o he blade, (D) s uc u e o he s ain-gauge senso s.
2.4. Da a P ocessing
The Sma Paddle calcula ion algo i hm has no been published. The Sma Paddle
gene a es he p ocessed da a ia he closed Ma lab GUI (G aphical Use In e ace) (Tam-
pe e, Finland) de eloped by he T ainesense Oy (Tampe e, Finland) meaning ha he e is
no access o Sma Paddle aw da a and he algo i hm calcula ion cons an s can no be
adjus ed. This was also he eason why he compa isons be ween he s ain-gauge sha
and he Sma Paddle used a e aged a iables as i was no possible o pe o m he ac ual
s oke (Sma Paddle)-s oke (s ain-gauge sha ) synch oniza ion due o lack o he aw
da a om he Sma Paddle. In he p esen s udy, he Sma Paddle a iables we e no mal-
ized o paddle blade (B acsa polo kine ic max, 735 cm
2
) a ea and he da a we e p ocessed
sepa a ely o he igh and le blades.
The mean o ce was he a e age o al o ce o ha ime when he blade is in he
wa e . The maximum o ce was he highes o al (all di ec ions) o ce ha he Sma Paddle
egis e ed du ing he en i e pe o mance. The o ce p oduc ion ime was de ined as he
du a ion du ing which he blade p oduced o ce a mo e han 30% o he s oke maximum
o ce. The absolu e o al impulse was o al o ce mul iplied by he whole ime when he
blade was unde he wa e su ace [35]. The Sma Paddle analysis consis ed o only he
c uising phase o he pe o mances wi hou he accele a ion and b aking phases. All a -
iables excep maximum o ce we e a e ages o he s okes included in he analysis. The
Figu e 1.
(
A
) Sma Paddle place on he blade, (
B
) S ain-gauge sha in calib a ion boa d (wi hou
IMU a ached), (C) Sma Paddle a ached o he blade, (D) s uc u e o he s ain-gauge senso s.
A s ain-gauge paddle sha (sel -manu ac u ed, Uni e si y o Jy äskylä, Finland) [
9
]
was used o he compa ison o he Sma Paddle (Figu e 1D). This same s ain-gauge
equipmen has also been used in p e ious esea ch when examining he ela ionships

Senso s 2022,22, 938 5 o 12
be ween o ce- ime pa ame e s and kayak eloci y [
9
]. The da a p oduced by he s ain-
gauge senso s we e ampli ied wi h sepa a e ampli ie s (TR es i Oy, Jy äskylä Finland) and
s o ed on a eco dable AD con e e (sel -manu ac u ed “LIIKE ca d”, Spo s Technology,
Uni e si y o Jy äskylä, Vuoka i, Finland) a a equency o 250 Hz. The gauge paddle
sha was calib a ed using 0 kg, 4 kg, and 10 kg weigh s o calcula e he linea gain ac o .
Du ing he calib a ion p ocess, he s ain gauge sha was a ached o he calib a ion able
wi h he clamps and he ulc um poin was a he edge o he able (Figu e 1B). The linea
gain ac o de e mined om he calib a ion was en e ed in o he eco dable AD Con e e
analysis so wa e.
2.4. Da a P ocessing
The Sma Paddle calcula ion algo i hm has no been published. The Sma Paddle
gene a es he p ocessed da a ia he closed Ma lab GUI (G aphical Use In e ace) (Tampe e,
Finland) de eloped by he T ainesense Oy (Tampe e, Finland) meaning ha he e is no
access o Sma Paddle aw da a and he algo i hm calcula ion cons an s can no be adjus ed.
This was also he eason why he compa isons be ween he s ain-gauge sha and he
Sma Paddle used a e aged a iables as i was no possible o pe o m he ac ual s oke
(Sma Paddle)-s oke (s ain-gauge sha ) synch oniza ion due o lack o he aw da a
om he Sma Paddle. In he p esen s udy, he Sma Paddle a iables we e no malized
o paddle blade (B acsa polo kine ic max, 735 cm
2
) a ea and he da a we e p ocessed
sepa a ely o he igh and le blades.
The mean o ce was he a e age o al o ce o ha ime when he blade is in he
wa e . The maximum o ce was he highes o al (all di ec ions) o ce ha he Sma Paddle
egis e ed du ing he en i e pe o mance. The o ce p oduc ion ime was de ined as
he du a ion du ing which he blade p oduced o ce a mo e han 30% o he s oke
maximum o ce. The absolu e o al impulse was o al o ce mul iplied by he whole ime
when he blade was unde he wa e su ace [
35
]. The Sma Paddle analysis consis ed o
only he c uising phase o he pe o mances wi hou he accele a ion and b aking phases.
All a iables excep maximum o ce we e a e ages o he s okes included in he analysis.
The ela i e o wa d, la e al and e ical impulses es ima ed by he Sma Paddle we e
calcula ed in line wi h kayaking di ec ion.
The da a s o ed om he s ain-gauge paddle sha was p ocessed using Ma lab
(R2020b, The Ma hWo ks Inc., Na ick, MA, USA). Fi s , he aw da a we e il e ed us-
ing a mo ing-a e age il e (mo mean unc ion) wi h a alue o 10 da a poin s ( il e
window) [
36
], o a enua e he la ges in e e ence peaks. Then he maximum o ce (max–
unc ion) o he paddle s oke was de e mined om he il e ed da a by iden i ying a
peak o ce alue (highes o ce) o he en i e pe o mance (c uising phase). Finally, mean
s oke o ce and o ce p oduc ion ime we e calcula ed using he same 30% o ce limi as
he Sma Paddle algo i hm in o de o make he analysis compa able. This was done by
de e mining a le el o 30% o he maximum o ce. The du a ion and mean o he o ce
p oduc ion ime we e de e mined om he o ce ec o abo e his le el. The le and igh
sides we e p ocessed sepa a ely in he analysis, wi h bo h sides o ming hei own da a
poin s in he s a is ical analysis. Mean alues (excep maximum o ce) o o al kayaking
pe o mances (c uising phase) we e used in his s ain-gauge paddle sha analysis o make
i compa able wi h he Sma Paddle analysis.
2.5. S a is ical Analysis
Resul s a e epo ed as mean and s anda d de ia ion (SD). Shapi o-Wilk no mali y
es and his og ams isual e alua ion (no mal dis ibu ion cu e and dis ibu ion skewness
and ku osis) we e used o check he no mali y o he da a. The Shapi o-Wilk es indica ed
ha some o he a iables we e no no mally dis ibu ed (s oke a e, boa eloci y, la e al
and e ical ela i e impulse). Simila ly, he isual in e p e a ion o he his og ams and
he small sample size (n=6&n= 14) suppo ed he selec ion o non-pa ame ic es s.
Associa ions be ween he da a measu ed by he Sma Paddle and he s ain-gauge paddle
Senso s 2022,22, 938 6 o 12
sha we e es ed wi h he Spea man co ela ion coe icien . The co espondence be ween
he da a measu ed by he Sma Paddle and he gauge-paddle sha was e alua ed wi h
wo-way andom (absolu e ag eemen ), single measu e in a-class co ela ion coe icien s
(ICC) [
37
], and isualized wi h Bland-Al man plo s [
38
]. ICC was used o cha ac e ize
he co espondence as poo (<0.40), ai (0.40 o <0.60), good (0.60 o <0.75), o excellen
(
≥
0.75) [
39
]. The ela ionship be ween eloci y and key paddle a iables measu ed wi h
he Sma Paddle was also s udied by he Spea man co ela ion coe icien s ( wo- ailed).
Di e ences be ween eloci ies g oups and s ain-gauge paddle sha and he Sma Paddle
we e es ed using ela ed–samples Wilcoxon signed- ank es . S a is ical signi icance was
se a p
≤
0.05 ( wo- ailed) and analyses we e pe o med wi h he R s a is ical compu ing
so wa e ( e sion 4.0.3) [40].
3. Resul s
3.1. S ain-Gauge Sha Compa ison
The Sma Paddle measu ed highe maximum o ce alues han he s ain-gauge
paddle sha (p< 0.001) (Table 2). The s ain-gauge paddle sha maximum o ce was
125.4 ±34.2 N
, while he mean o he maximum o ce o he Sma Paddle was
152.1 ±57.4 N
.
Howe e , he co ela ion be ween Sma Paddle and s ain-gauge paddle sha was s ong
0.84 (p< 0.001), and he ICC was classi ied o be good 0.64 (Con idence In e al CI 95%
0.21, 0.82) (p= 0.003). The s ain-gauge paddle sha mean o ce was 85.4
±
25.3 N and he
Sma Paddle 83.3
±
30.3 N, wi h a s ong co ela ion o 0.88 (p< 0.001) and an excellen
ICC o 0.86 (CI, 0.78, 0.92) (p< 0.001). This di e ence in mean o ce was no s a is ically
signi ican be ween hese wo me hods (p= 0.083).
Table 2.
Compa ison be ween he s ain-gauge paddle sha and he Sma Paddle (9-axis IMU+
p essu e senso ) in key paddle s oke a iables (n= 6).
Mean (SD) S ain-Gauge Sma Paddle p-Value 1 2ICC 3
Maximal o ce [N] 125.4 ±34.2 152.1 ±57.4 <0.001 0.84 ** 0.64 ***
Mean o ce [N] 85.4 ±25.3 83.3 ±30.3 =0.083 0.88 ** 0.86 ***
Fo ce p od. ime [s] 0.43 ±0.08 0.44 ±0.10 =0.472 0.88 ** 0.81 ***
S oke a e [1/min] 84.8 ±25.6 80.7 ±21.0 =0.199 0.95 ** 0.95 ***
SD = S anda d de ia ion,
1
Wilcoxon signed ank es ,
2
Spea man’s ank co ela ion coe icien ,
3
ICC = In aclass
Co ela ion Coe icien , ** Co ela ion is s a is ically signi ican 0.01 le el (2- ailed), *** p< 0.001.
The o ce p oduc ion ime s ongly co ela ed be ween he s ain-gauge paddle sha
and he Sma Paddle (0.88, p< 0.01) and he a e age di e ence was only 0.01 s (p= 0.472)
The ICC was classi ied o be excellen 0.81 (CI, 0.70, 0.88) (p< 0.001). O all he a iables,
he s onges co ela ion was ound in he s oke a e 0.95 (p< 0.01) and i did no di e
s a is ically (p= 0.199), and he ICC was excellen a 0.95 (CI, 0.89, 0.98) (p< 0.001) (Table 2).
Sca e plo s and R
2
alues be ween Sma Paddle and s ain-gauge sha in key paddle
s oke a iables a e p esen ed in Figu e 2.
The Bland-Al man analysis (Figu e 3) showed ha he e we e only a ew cases ou side
he conco dance limi s (95%) and no sys ema ic e o di e ence was no iceable, as he
measu emen esul s we e e enly dis ibu ed on bo h sides o he ze o le el.
3.2. Key Va iables o Veloci y
S oke a e ( = 0.86, p< 0.001) and mean o ce ( = 0.79, p< 0.001) we e mos
s ongly ela ed o boa eloci y and di e ed be ween all kayaking eloci ies (p< 0.001)
(Tables 3and 4)
. The absolu e impulse ( = 0.70, p< 0.001) was also posi i ely ela ed o
he eloci y o he boa and di e ed signi ican ly (p< 0.001) be ween he eloci ies while
he o ce p oduc ion- ime was nega i ely ela ed ( =
−
0.76, p< 0.001) o he eloci y o
he boa .
Senso s 2022,22, 938 7 o 12
Senso s 2022, 22, x FOR PEER REVIEW 7 o 13
2). Sca e plo s and R2 alues be ween Sma Paddle and s ain-gauge sha in key paddle
s oke a iables a e p esen ed in Figu e 2.
Table 2. Compa ison be ween he s ain-gauge paddle sha and he Sma Paddle (9-axis IMU+
p essu e senso ) in key paddle s oke a iables (n = 6).
Mean (SD) S ain-Gauge Sma Paddle p-Value 1 2 ICC 3
Maximal o ce [N] 125.4 ± 34.2 152.1 ± 57.4 <0.001 0.84 ** 0.64 ***
Mean o ce [N] 85.4 ± 25.3 83.3 ± 30.3 =0.083 0.88 ** 0.86 ***
Fo ce p od. ime [s] 0.43 ± 0.08 0.44 ± 0.10 =0.472 0.88 ** 0.81 ***
S oke a e [1/min] 84.8 ± 25.6 80.7 ± 21.0 =0.199 0.95 ** 0.95 ***
SD = S anda d de ia ion, 1 Wilcoxon signed ank es , 2 Spea man’s ank co ela ion coe icien , 3
ICC = In aclass Co ela ion Coe icien , ** Co ela ion is s a is ically signi ican 0.01 le el (2-
ailed), *** p < 0.001.
Figu e 2. Sca e plo s and R2 alues be ween Sma Paddle and s ain-gauge sha in key paddle
s oke a iables (n = 6) (R2 = coe icien o de e mina ion showing how close he da a a e o he i ed
eg ession line).
The Bland-Al man analysis (Figu e 3) showed ha he e we e only a ew cases ou -
side he conco dance limi s (95%) and no sys ema ic e o di e ence was no iceable, as
he measu emen esul s we e e enly dis ibu ed on bo h sides o he ze o le el.
Figu e 2.
Sca e plo s and R
2
alues be ween Sma Paddle and s ain-gauge sha in key paddle
s oke a iables (n= 6) (R
2
= coe icien o de e mina ion showing how close he da a a e o he i ed
eg ession line).
Senso s 2022, 22, x FOR PEER REVIEW 8 o 13
Figu e 3. Bland–Al man analysis be ween Sma Paddle and s ain-gauge sha in key paddle s oke
a iables (n = 6).
3.2. Key Va iables o Veloci y
S oke a e ( = 0.86, p < 0.001) and mean o ce ( = 0.79, p < 0.001) we e mos s ongly
ela ed o boa eloci y and di e ed be ween all kayaking eloci ies (p < 0.001) (Tables 3
and 4). The absolu e impulse ( = 0.70, p < 0.001) was also posi i ely ela ed o he eloci y
o he boa and di e ed signi ican ly (p < 0.001) be ween he eloci ies while he o ce
p oduc ion- ime was nega i ely ela ed ( = −0.76, p < 0.001) o he eloci y o he boa .
Table 3. Desc ip i e cha ac e is ics o he pa icipan s and esul s o he es ing eloci ies (n = 14).
Mean (SD) Veloci y 1 Veloci y 2 Di
V1–V2 1 Veloci y 3 Di
V2–V3 1
Veloci y [m/s] 2.1 ± 0.2 2.4 ± 0.1 <0.001 2.7 ± 0.3 <0.001
Maximal o ce [N] 112.6 ± 38.4 145.1 ± 40.1 <0.001 224.9 ± 67.8 <0.001
Mean o ce [N] 61.6 ± 17.9 83.6 ± 21.8 <0.001 115.8 ± 32.2 <0.001
Fo ce p od. ime [s] 0.46 ± 0.06 0.40 ± 0.06 <0.001 0.34 ± 0.07 <0.001
S oke a e [1/min] 75.7 ± 14.3 93.5 ± 13.4 <0.001 119.3 ± 19.7 <0.001
To al impulse abs [Ns] 58.6 ± 16.8 69.2 ± 15.8 <0.001 90.1 ± 22.8 <0.001
Impulse o wa d [%] 2 49.50 ± 4.29 50.36 ± 4.23 0.387 49.75 ± 4.05 0.770
Impulse la e al [%] 2 28.81 ± 4.90 29.60 ± 5.10 0.058 32.08 ± 5.14 0.036
Impulse e ical [%] 2 21.69 ± 4.45 20.04 ± 3.81 0.002 18.17 ± 4.24 0.009
SD = S anda d de ia ion, 1 Rela ed-samples Wilcoxon signed- ank es , 2 Po ion o o al impulse.
The ela i e la e al impulse was posi i ely and mode a ely ela ed o boa eloci y (
= 0.42, p < 0.001) while he e ical impulse was nega i ely and mode a ely ela ed o e-
loci y ( = −0.42, p < 0.001). The e was no change in he o wa d impulse as he eloci y
inc eased and i did no co ela e signi ican ly wi h he eloci y ( = −0.09, p = 0.429).
Figu e 3.
Bland–Al man analysis be ween Sma Paddle and s ain-gauge sha in key paddle s oke
a iables (n= 6).
Senso s 2022,22, 938 8 o 12
Table 3. Desc ip i e cha ac e is ics o he pa icipan s and esul s o he es ing eloci ies (n= 14).
Mean (SD) Veloci y 1 Veloci y 2 Di
V1–V2 1Veloci y 3 Di
V2–V3 1
Veloci y [m/s] 2.1 ±0.2 2.4 ±0.1 <0.001 2.7 ±0.3 <0.001
Maximal o ce [N] 112.6 ±38.4 145.1 ±40.1 <0.001 224.9 ±67.8 <0.001
Mean o ce [N] 61.6 ±17.9 83.6 ±21.8 <0.001 115.8 ±32.2 <0.001
Fo ce p od. ime [s] 0.46 ±0.06 0.40 ±0.06 <0.001 0.34 ±0.07 <0.001
S oke a e [1/min] 75.7 ±14.3 93.5 ±13.4 <0.001 119.3 ±19.7 <0.001
To al impulse abs [Ns] 58.6 ±16.8 69.2 ±15.8 <0.001 90.1 ±22.8 <0.001
Impulse o wa d [%] 249.50 ±4.29 50.36 ±4.23 0.387 49.75 ±4.05 0.770
Impulse la e al [%] 228.81 ±4.90 29.60 ±5.10 0.058 32.08 ±5.14 0.036
Impulse e ical [%] 221.69 ±4.45 20.04 ±3.81 0.002 18.17 ±4.24 0.009
SD = S anda d de ia ion, 1Rela ed-samples Wilcoxon signed- ank es , 2Po ion o o al impulse.
Table 4. The Spea man’s ank co ela ion coe icien by eloci y g oups (n= 14).
Veloci y 1 Veloci y 2 Veloci y 3 All Veloci y
Maximal o ce [N] 0.60 ** 0.56 ** 0.74 ** 0.78 **
Mean o ce [N] 0.61 ** 0.47 * 0.73 ** 0.79 **
Fo ce p od. ime [s] −0.54 ** −0.40 * −0.56 ** −0.76 **
S oke a e [1/min] 0.82 ** 0.34 0.76 ** 0.86 **
To al impulse abs [Ns]
0.43 * 0.52 ** 0.74 ** 0.70 **
Impulse o wa d [%] −0.15 −0.01 −0.18 −0.09
Impulse la e al [%] 0.45 * 0.11 0.58 ** 0.42 **
Impulse e ical [%] −0.27 −0.07 −0.46 * −0.42 **
*p< 0.05, ** p< 0.01 ( wo- ailed).
The ela i e la e al impulse was posi i ely and mode a ely ela ed o boa eloci y
( = 0.42, p< 0.001) while he e ical impulse was nega i ely and mode a ely ela ed o
eloci y ( =
−
0.42, p< 0.001). The e was no change in he o wa d impulse as he eloci y
inc eased and i did no co ela e signi ican ly wi h he eloci y ( = −0.09, p= 0.429).
4. Discussion
The p esen s udy sugges s ha he T ainesense Sma Paddle
®
o e s p omising op-
po uni ies o measu e paddle s oke key a iables acknowledging ha he calcula ion
algo i hms and de ice moun ing need o be u he de eloped mo e sui able o kayaking
in he u u e. The s oke a e, o ce, o ce- ime, and o al absolu e impulse we e ound
o be key a iables a ec ing he eloci y o he boa among canoe polo playe s and hese
esul s we e also well in line wi h p e ious s udies wi h sp in kayake s [
5
,
8
]. Mo eo e , he
T ainesense Sma Paddle
®
made i possible o s udy he di ec ion o he s oke, and in his
s udy, when he eloci y o he boa inc eased, he ela i e amoun o he la e al impulse o
he paddle s oke inc eased, while he e ical dec eased. The T ainesense Sma Paddle
can p o ide new insigh s o echnical coaching in kayaking in eal open wa e condi ions.
In his s udy, good o excellen ag eemen was ound be ween he Sma Paddle and
he s ain-gauge sha in key a iables o paddle s oke (0.64–0.95, p< 0.001). The absolu e
di e ences be ween he Sma Paddle and he e e ence me hod in o ce p oduc ion ime,
mean o ce, and s oke a e we e small. The Sma Paddle o e es ima ed maximum o ce
compa ed o a s ain-gauge sha (p> 0.001) and he ag eemen was also lowes in his
a iable (0.64, p< 0.001). This may be due o di e ences in da a p ocessing, especially da a
il e ing. We used a mo ing a e age (mo mean) il e o analyze he s ain-gauge sha
da a, which may cu he highes o ce peaks mo e han he il e ing algo i hm used by he
Sma Paddle. Howe e , i was mo e likely ha he Sma Paddle o e es ima ed he highes
alues such as maximum o ce and absolu e impulse due o issues ela ed o hei ha dwa e
and algo i hms. This was indica ed by he sensi i i y analysis be ween he le and igh
sides which sugges ed ha i was indeed he Sma Paddle ha o e es ima ed he maximum
o ce compa ed o he s ain-gauge sha (Supplemen a y Table S1). Un o una ely, his