* Co esponding au ho : imo eie [email protected]
Redesigning he g ipping sys em o KUKA KR5 by u ilizing 3
dimension p in ing echnologies and A duino
Vik o Va ga1, Timo ei Is án E dei1,*, and Géza Husi1
1Uni e si y o Deb ecen, Mecha onics Depa men , 4028 Ó eme ő s ee 2. Deb ecen, Hunga y
Abs ac . Du ing his sho summa y o ou p ojec , We’d like o go h ough all he s eps o he p ojec ,
om he e y i s planning s ages, up o he 3D modelling, and inishing up wi h he ac ual p in ing i sel .
We’ e also had he chance o ins all ou inished ool on o he KUKA KR5 i sel , and e alua e he quali y o
ou ool using senso s, measu emen , and by analysing ou A duino code.
1 In oduc ion
Du ing ou cou se in Robo ics, we had he chance o
s udy he KUKA KR5 obo , wi h many o he obo s [1]
and au oma ed sys ems [2], bo h heo e ically, and
p ac ically, om he basic oolse o he obo ’s inne
code and s uc u e. A he momen , he obo uses a wo
inge pneuma ic g ippe as i s ool. This g ippe can
hold objec s anging o 30 o 45 [mm] in diame e , and
cu en ly used o g ip on a ool used o educa ional
pu poses. This me hod is su icien , bu a om op imal,
conside ing i s small dynamic ange, wo-s a e (open o
closed) na u e only enables small objec s o wo k wi h,
and mo eo e hey canno be agile as a esul o ei he
maximum o ze o g ipping s eng h. Because he Lab
does no possess any mo e g ipping ools, i was an
ob ious ini ia i e o c ea e mo e a ied g ippe bo h o
inc ease he a ie y o ools, and o challenge us wi h
his endea ou . We choose a 3 inge pa allel g ippe . As
a esul o he sp ead o 3D p in ing echnologies, we
could physically mani es he ool in eal li e as well.
The supe ision and con ol has been done wi h a
Blue oo h module using an A duino. Because he inge s
a e unde cons an su eillance by he senso , we can ge
accu a e measu emen s on he objec held on o on he
ly. We used a FSR (Fo ce Sensi i e Resis o )
dynamome e o ge he cu en clamping o ce on he
inge s. A e in eg a ing he ool in o he es
en i onmen , we p oceeded o ca y ou some
measu emen s and es ing, o make su e ou p ojec was
success ul.
2 P esen ing he mechanism
As ou i s s ep, be o e we s a ed he 3D modelling, we
had o speci y he me hod o mo ing he inge s. The
inge s a e 120° o m one ano he , and a e being mo ed
by a o a y componen (Fig. 1., Fig. 2.).
Fig. 1. P esen ing he mechanism
This enabled us o con e o a ional mo emen o
linea . Same me hod was used in s eam engines, o being
used in mode n ecip oca ing (pis on) engines,
comp esso s o pumps.
Nex s ep was o de ine he g ea es angle which
g an s us sa e and accu a e, collision ee mo emen .
A e he measu emen s, we go he p ecise angle o
106° o be used as he maximum ange, which enables us
o ha e a scope o 50 [mm] o g ipping, which p o ed
o be adequa e. By knowing hese alues, we can use a
ske ch d awn in Fusion 360 [3] o accu a ely measu e he
upcoming componen s’ sizes, and also o keep ack o
he angles, and mechanical model o ou sys em. On he
le side o he igu e, we can see he mo o a 0°, hus
i ’s ully open posi ion, while on he igh we can see
maximal 106°, app oxima ely 50 [mm] long con ac ion.
Fig. 2. P esen ing he mechanism
ANNALS OF THE UNIVERSITY OF ORADEA
Fascicle o Managemen and Technological Enginee ing
ISSUE #2, AUGUST 2018, h p://www.im uo adea. o/auo. m e/
93
3 P esen ing he 3D model
A e we success ully d awn ou ou pa s, and measu ed
hem co ec ly, we can ge o 3D modelling he pa s o
ou g ippe . We used Au odesk Fusion 360 o c ea e ou
3D models. In he upcoming igu es, we can see he pa s
o ou g ippe sepa a ely. On Fig. 3. we can see he
ame o he g ippe , which houses ou mo o , and he
ci cula slide s placed 120° apa , o enable he inge s o
mo e.
Fig. 3. F ame
The images down below (Fig. 4.) a e he al eady
p esen ed 44 [mm] long push ods (le ), and he o a y
componen moun ed o he mo o la e ( igh ).
Fig. 4. Push od and o a y componen
Fig. 5. shows ou 3D modelled inge , which is made
so ha i can slide in he ci cula slide o he ame
e o lessly.
Fig. 5. Push od and o a y componen
Ano he ask was o c ea e co ec a i ing
mechanism in be ween he obo and he ool i sel . The
6 h axis o he KUKA KR5 possesses 4 moun ing poin s
whe e we can a ach ools. By knowing he exac
posi ion and size o hese bo es, we made an adap e
(Fig. 6.) so we can i he ool on o he obo easily.
Fig. 6. The adap e
4 Assembling he pa s in Fusion 360
As he i s s ep o he assembly, we connec ed he
Towe P o MG995 se o o ou ame, hen we moun ed
he o a y componen on o i . The chosen se o has g ea
ange and g ea o que compa ed o i s size. The able
below shows some echnical da a abou i :
Table 1. Speci ica ions o he se o [4]
Se o ype
To que [Nm]
Speed
[s/60°]
Weigh
[g]
Towe P o
MG995
1
0,16
55
Coming up, we added he inge s, sled in om he
side in o he ame’s slide s, connec ed wi h he
push ods o which he ea e connec ed o he o a y
componen . This concludes ou 3D model o he g ippe
(Fig. 7.).
Fig. 7. Assembled 3D model
The g ip ange is he di e ence in be ween he ully
open and closed posi ions, which le s us g ab objec s
anging om 9 [mm] o 109 [mm]. The adap e and he
g ippe i sel can be sled oge he , and hen secu ed
oge he wi h sc ews (Fig. 8.).
ANNALS OF THE UNIVERSITY OF ORADEA
Fascicle o Managemen and Technological Enginee ing
ISSUE #2, AUGUST 2018, h p://www.im uo adea. o/auo. m e/
94
Fig. 8. Joining he adap e and he g ippe oge he
5 P in ing ou he componen s in 3D
The p in ing has been done by a hi d pa y, a e we
ag eed in he e ms and ma e ials o he p in (Fig. 9.).
Gene ally, ABS is mo e duc ile, hea esis an and less
p one o sha e ing, bu since we won’ need hese
ad an ages, since we won’ mo e any hing hea y o ho ,
PLA was a cheape al e na i e. I ’s also p e e ed by
hobbyis s and designe s also endo se i o i s ease o use
and la ge colou a ia ion [5].
Fig. 9. The componen s p in ed om PLA
Ul imake Cu a akes he bu den o se ing all he
pa ame e s o each p in om us, since i au oma ically
se s i o us, sa ing a lo o ime. The de ails can be
ound in he nex able [6].
Table 2. P in ing de ails pe componen s
Componen
Fill
ac o
[%]
To al p in
ime
[hou :min]
Weigh
[g]
Cos [F ]
1
30
2:48
31,1
165
2
30
6:54
82,1
435,1
3 (3 pieces)
30
2:54
23,7
126,6
4
30
00:10
1,6
8,5
5 (3 pieces)
30
00:06
1,2
6,3
Σ
N/A
12:52
139,7
741,5
The o al weigh o he g ippe wi hou he sc ews is
only 194,7 [g], which akes up he 3,89% o he
maximum o 5 [kg] load on he obo ’s las axis. A e
we manu ac u ed he pa s, we assembled in he same
way as we planned i ou in he i ual wo ld (Fig. 10.).
To u he inc ease he g ip s eng h, we used ubbe
ends on he inge ips.
Fig. 10. The assembled g ippe
6 Con olling he g ippe wi h A duino
As he basic me hodology o se os, e en i hey hi an
obs acle while u ning, hey won’ s op, qui e he
con a y, hey will y o con inuously u n un il he
po en iome e inside eaches he gi en angle [7]. This
can cause se ious o e hea ing, and la e he ailu e o he
se o. To p ojec his issue on o ou design, i s no
p e e able o d i e he se o o 0° o 106° as a me hod o
g ipping, since i will kill he se o in a sho ime. Thus,
we ied o achie e he op imal s a e o g ipping on he
objec , hen locking he se o a ha angle, p e en ing
u he mo emen s, bu also g ipping on i s eadily.
The issue was sol ed by analysing he po en iome e
inside he se o. A e we solde ed a pin om he
A duino o he signal line o he po en iome e , we
could easily, and accu a ely measu e he s a e o he
se o a any ime in a scale o 0 o 1024 uni s. hen we
used command “Map” o map ou he po en iome e
alues o each angle, 122 o 20° and 302 o 126° ( his
+20 shi will be bene icial la e ).We sa ed his o he
al a iable, which will now always show +20 om he
eal alue, which should be accoun ed o .
The p og am seen on Fig. 11. basic p emise is o
d i e o mo o om 0° o 106° inside a loop,
inc emen ing he angle in each loop. As s a ed abo e, he
al in ege will keep inc easing wi h he loop, as he
mo o is u ning abou , bu i i ge s s uck, i will cease o
inc ease. As he g ip happened, he loop will con inue o
un, and as I equals wi h he al in ege , we can sa ely
say ha he g ip has happened, so we se he se o o he
loop’s local a iable x-17. The sub ac ion impo an ,
since as al equals o x we a e al eady o e u ning by
20°, so we s ep ha down o only 3°which p oduces a
i m g ip, ye i s ligh on he se o.
ANNALS OF THE UNIVERSITY OF ORADEA
Fascicle o Managemen and Technological Enginee ing
ISSUE #2, AUGUST 2018, h p://www.im uo adea. o/auo. m e/
95
Fig. 11. A snippe om he A duino code
The g ippe can be con olled ia Blue oo h, wi h he
Module HC-06, along wi h a ee And oid app
downloadable online [8]. The App can send wo
commands o he Module, on he i s p ess, i s a s he
con ac ion, by sending a + signal, and hen we can p ess
i again o e e se he con ac ion, loosening i , wi h he -
signal. We’ e d awn he wi ing diag am in F i zing,
shown on Fig. 12. [9][10].
Fig. 12. The A duino wi ing diag am
The powe o he mo o is supplied by ex e nal
sou ce, since he A duino canno gi e i enough cu en
o un smoo hly. We also had o wi e in a 1 [kΩ] esis o ,
so he signals would no luc ua e so much, and we
g ounded he whole con ap ion o a common g ound so
we achie e he desi ed esul s.
7 Calcula ing he diame e o he held
objec
By s opping he se o on he g ip posi ion, we can
p ecisely calcula e he diame e o he objec held on o,
using he posi ion o he se o. The code p esen ed i is
ob ious, ha we can use he mo o ’s p esen angle s o ed
in x o calcula e he diame e o said objec . We will go
h ough he calcula ion in a pa ame ic ashion. On he
igu e below (Fig. 13.) we can see he symbols, and o
added me hodical app oach, we added he indica ions o
bo h le and igh igu es.
Fig. 13. No able poin s and edges o he mechanical model
(1)
As we’ e al eady s a ed x is always shows 20° mo e
han he eal alue, so we sub ac i .
(2)
(3)
(4)
(5)
(6)
(7)
(8)
L1 and L2 a e cons an s.
The 8 h o mula ep esen s he 20 [mm] shi om
he calcula ed o he eal posi ion. The mul iplica ion o
he adius wi h 2 equa es o he diame e o he objec .
A e we decla ed he necessa y a iables, we simply
implemen ed he o mula in o he A duino code (Fig.
14.), some ime consolida ed oge he o sa ing space
(and memo y).
Fig. 14. Calcula ing he diame e o he held objec
8 De e mining he clamping o ce using
he FSR 400 dynamome e
We’ e chose he FSR 400 (Fo ce Sensi i e Resis o )
dynamome e o his ask. The senso is basically a
s ain gauging esis o inc easing i s esis ance wi h
p essu e in be ween 0,2 [N] and 20 [N] [11]. The
measu emen can be ca ied ou easily, isible on his
simple wi ing diag am:
ANNALS OF THE UNIVERSITY OF ORADEA
Fascicle o Managemen and Technological Enginee ing
ISSUE #2, AUGUST 2018, h p://www.im uo adea. o/auo. m e/
96
Fig. 15. Wi ing diag am o he FSR 400 measu emen [12]
One pin o he senso goes o he 5 [V] powe supply,
he o he goes o he common g ound a e a sc ape
esis ance. Addi ionally, we connec ed he analog pin in
be ween he sc ape esis ance and he FSR 400 (Fig.
15.). A e o e coming he minimal h eshold load
(p essu e), he FSR ac s like a esis o o 100 [kΩ].
P opo ionally wi h he p essu e, he esis ance d ops
om 100 [kΩ] o a 10 [kΩ] minimum. This means an
inc easing ol age on he line, equa ing o da a o us on
he analog pin. A e he wi ing had been done, we
c ea ed a p og am o acqui e he esul s om aw ol age
(Fig. 16.).
Fig. 16. The code o he FSR 400 dynamome e
Fi s ly, as usual, we decla ed ou a iables. The main
loop i s p in s he aw alue o he analog pin on o he
se ial plo e . This equa es om 0 [V] o 5 [V]
p opo ional o he p essu e, o inc ease he esolu ion,
we used he command map o con e his om 0 [mV]
o 5000 [mV]. We p in ed ou he ol age o he se ial
moni o as well, in milli ol s. Wi hou any load, he
p og am simply w i es ou : “No P essu e”. We can
calcula e he ol age on he sc ape esis ance wi h he
ollowing o mula:
(9)
By so ing he equa ion, we can ge he o mula:
(10)
U = 5 [V], R = 10 [kΩ]
We implemen ed his o mula in o ou A duino code,
esul ing he FSR senso ’s esis ance in [Ω] (Fig. 17.).
To measu e he clamping o o he g ippe , we use a 30
[mm], a 60 [mm] and a 90 [mm] diame e es p op,
p in ed ou in 3D [13]. We secu ed each p op wi h he
g ippe , while ha ing he FSR 400 on one o he inge s.
We can see he esul s o he 60 [mm] objec on he
image below.
Fig. 17. Measu emen esul s o he FSR 400
On he le side, he al eady p esen ed diame e
calcula ion is isible, and on he igh , we can see he
measu emen esul s he FSR esis ance’s esul s a e
necessa y o measu e he o ce ac ing upon he senso ,
isible on he diag am below, whe e we display he
esul s on he y axis, hen in e sec ing ha a 30,60 o 90
(diame e ), hen in e sec ing hose poin s om he a c o
he x axis, hen di ide hese alues by 100, we can ge
he co ec o ce in [N].
Fig. 18. Diag am o he FSR dynamome e
ANNALS OF THE UNIVERSITY OF ORADEA
Fascicle o Managemen and Technological Enginee ing
ISSUE #2, AUGUST 2018, h p://www.im uo adea. o/auo. m e/
97

In e p e ing he alues, we can see ha he inge ac s
wi h 1,7 [N], 1,9 [N] and 2,05 [N] on di e en diame e s
(Fig. 18.).
9 Re e ence wo k
A e comple ing he se up, we ins alled he g ippe wi h
he adap e men ioned abo e o he obo (Fig. 19.), also,
we igged he A duino on o he obo as well, o
simplici y’s sake.
Fig. 19. Rigging he g ippe on o he obo
A e ins alling, we manipula ed di e en objec s,
li ing hem, o g ipping hem alike (Fig. 20.).
Fig. 20. The wo king obo
9 Conclusion
We can con iden ly say ha we c ea ed a wo king
g ipping ool o he KUKA KR5, and acqui ed use ul
and p ac ical knowledge on he way. The code and
senso wo king oge he can measu e diame e and
clamping o ce, adding ano he laye o complexi y o
ou sys em. In his pape we wen o e he de ailed
c ea ion o he whole p ojec , and we a e conside ing
u he imp o emen s o he ool.
Acknowledgmen
The publica ion was suppo ed by he p ojec EFOP-
3.6.1-16-2016-00022. This p ojec was co- unded by he
Eu opean Union, om he Eu opean Social Fund.
Re e ences
1. Zs. Molná , T. I. E dei, N. C. Obinna, G. Husi,
„A no el design o an ai -cushion ehicle and
i s implemen a ion,” MATEC Web Con .
Volume 126
2. T. I. E dei, Zs. Molná , N. C. Obinna, G. Husi,
„Su eillance and Secu i y Sys em in he
Building Mecha onics Resea ch Cen e ,”
Elec ical Enginee ing and Mecha onics
Con e ence EEMC’16 - 7 h-18 h-19 h Ma ch,
2016.
3. „Au odesk Fusion 360,” [Online]. A ailable:
h ps://www.au odesk.com/p oduc s/ usion-
360/o e iew. [Access Da e: 27 11 2017].
4. „Towe P o MG995 Se o,” [Online]. A ailable:
h ps://se oda abase.com/se o/ owe p o/mg99
5. [Access Da e: 27 11 2017].
5. „ABS o PLA: Which 3D p in ing ilamen
should you use?,” [Online]. A ailable:
h ps://www.digi al ends.com/cool- ech/abs- s-
pla-3d-p in ing-ma e ials-compa ison/. [Access
Da e: 27 11 2017].
6. „Ul imake Cu a so wa e,” [Online].
A ailable:
h ps://ul imake .com/en/p oduc s/ul imake -
cu a-so wa e. [Access Da e: 27 11 2017].
7. F. Reed, „How Do Se o Mo o s Wo k,”
[Online]. A ailable:
h ps://www.jameco.com/jameco/wo kshop/ho
wi wo ks/how-se o-mo o s-wo k.h ml.
[Access Da e: 27 11 2017].
8. „A duino Blue oo h con olle ,” [Online].
A ailable:
h ps://play.google.com/s o e/apps/de ails?id=c
om.giumig.apps.blue oo hse ialmoni o .
[Access Da e: 27 11 2017].
9. „A duino,” [Online]. A ailable:
h ps://www.a duino.cc/. [Access Da e: 27 11
2017].
10. „F i zing,” [Online]. A ailable:
h p:// i zing.o g/home/. [Access Da e: 27 11
2017].
11. „Fo ce Sensi i e Resis o (FSR), O e iew,”
[Online]. A ailable:
h ps://lea n.ada ui .com/ o ce-sensi i e-
esis o - s /o e iew. [Access Da e: 27 11
2017].
12. „Fo ce Sensi i e Resis o (FSR), Using an
FSR,” [Online]. A ailable:
h ps://lea n.ada ui .com/ o ce-sensi i e-
esis o - s /using-an- s . [Access Da e: 27 11
2017].
13. T. I. E dei, Zs. Molná , N. C. Obinna, G. Husi,
„A No el Design o an Augmen ed Reali y
Based Na iga ion Sys em & i s Indus ial
Applica ions,” 15 h IMEKO TC10 – Technical
Diagnos ics in Cybe -Physical E a Budapes ,
Hunga y, 6 – 7 June, 2017 - O ganised by:
MTA SZTAKI – Hunga ian Academy o
Sciences - Ins i u e o Compu e Science and
Con ol.
ANNALS OF THE UNIVERSITY OF ORADEA
Fascicle o Managemen and Technological Enginee ing
ISSUE #2, AUGUST 2018, h p://www.im uo adea. o/auo. m e/
98