0 - The Robotics Institute - Carnegie Mellon University

A Functional Vehicle for
Autonomous Mobile Robot Research
Gregg Podnar, Kevin Dowling, and Mike Blackwell
CMU-RI-TR-84-28
Autonomous Mobile Robots Laboratory
The Robotics Institute
Carnegie-MellonUniversity
Pittsburgh, Pennsylvania 15213
April 1984
Copyright @ 1984 Carnegie-MellonUniversity
!
i
Abstract
1. Introduction
2. Ilcsign Kequirmncnts
3. Design Occisions
3.1 Chassis and Suspcnsion
3.2 l’cthcr
3.3 Powcr
3.4 Motors
3.5 Control
3.6 Communication
4. Structure
4.1 The Fork
4.2 The Frame
4.3 Construction
4.4 Phasc Shifting Components
4.5 Manual Control
4.6 1,imit Switches
4.7 Camera Mount
5. Performance
5.1 Mcchanical
5.2 Moving Accuracy
5.3 System Pcrfonnance
5.4 Summary
6. Elcctrical Control
6.1 System Power
6.2 Motor switching
6.3 Digital Control Logic
6.4 Dclay Circuit
6.5 Main Control Circuitry
6.6 Layout
6.7 Possiblc Enhanccmcnts
7. On-Board Computer
8. Bibliography
Appcndix I
Mechanical Drawings
Appendix I1
Elcctrical Drawings
Appcndix 111
On-Board Processor Software
V
1
3
5
5
5
5
5
6
6
9
9
10
11
12
12
12
13
15
15
15
17
17
19
19
20
20
21
21
23
23
25
27
ii
iii
L i s t of 1”g
‘ I urcs
1:igurc 2- 1 :
Iiigurc 3-I :
Yigurc 4- I:
Kgurc 4-2:
Iiigurt 4-3:
1:igurc 4-4:
Figure 4-5:
Iigurc 5- I:
Figure 6-1:
Jigurc 6-2:
Figorc 6-3:
Figure 7-1:
Niyruz:e
Ncpmtto - Cutaway
Nry/utrci - ‘l‘hcbasic vchiclc (port sidc vicw).
N q m i r a - Aft of port. (Countcrwcightsnot installed.)
I .iinil Switch Ihngcs
A’rplutic - Caincl-ciSupport
Nqmtie - l’hc first run
Nepfuric- N‘ivigating in the I-ab
Start/Stop System
1)clay Circuit
Caincra Circuitry
On-13oard Proccssor Mock Diagram
4
8
9
11
13
13
14
17
19
21
22
25
iv
Abstract
N r p f u m is a tcthcrcd \chick built for aiitonoitiws mokilc robot rcscarch. Incliidcd arc: tlic dcsign
considcrations, UIC resulting design, ; k i d dctails of thc mcclxinical structurc and clcctrical conti 01 systcm.
I)ct;iil is sufficicnl to cnahlc rcpliciititw or adaptation by othcrs. A discussion of ttw pcrformancc with rcspcct
to thc Jcsign coilsidcratioiis is alx) iilcluded.
Every instrument, tool, vessel,
if it does that for which it is made,
is well.
Marcus Aurclius
1
13asic rcscai~chin mobilc rohotics rcquircs rcnl-world niobilc vchiclcs without wliich ninny topics arc liinitcd
lo Lliought cxp."rirncnts. I n dcvc.loping any vchiclc, its apyliccrrirm arc tlic soiircc of dcsigii considcrations.
One of thc arcniis of study is Artificial Jntclligcncc (AI). Our application is in ~ i u i o i i o i m t . rmohilc robots, that
is, robots which can carry out tasks nith no human intervention, attaining knowlcdgc of thcir cnvironmcnt,
and intcracting with it, by using Artificial Iiitclligcncc, (AI).
A vchiclc with complcx control rcquircmcnts or inconsistcnt reliability can contrihutc incfficicncy lo the
cxpcrimcntnl prtxcss [2]. Anothcr, oftcn substantial portion of thc total timc av;iilablc for rcscarch is
consumcd in fiibrication and dcbugging thc vehicle. By default, a vchiclc itsclf, with its associated
subsystcms, can bccomc a rcscarch projcct within thc major objectives [3].
Robot vchiclcs and lhcir basic controls are certainly worthy topics for rcscarch in thcir own right[5].
Howcvcr, a scparation ofconccrns may contribute to a morc rapid advanccment of thc art. Ily minimihg the
'lower-lcvcl' subsystcrns of a vchicle, the functionality should bc more stable. 'l'hc use of such a vchiclc in
'highcr-level' rcscarch shoulJ rcsult in a less intcrrupted study.
For thc purposcs of some of our AI research, an uncomplicated hnctional mobilc base was dcsigncd and
built. Its rcliability, and simplicity of operation continue to facilitatc experimcntation in our lab.
2
3
2. Design Requirements
‘I‘hc fiindomcntal dcsign rcquircmcnt is: a sclf-propcllcd vchiclc, with provisions for computcr control of
dircction and rnotivatioii. ‘lhisinobilc basc should bc cayablc of carrying:
o camcras
and lights for vision rcscarch,
e infrarcd and Sonar scnsors for ranging and proximity rcscitrch,
0
on-board proccssing for local control,
0
communication for monitoring, rcmotc control, and off-board proccssing,
e manipulators
for actively modifying thc cnvironmcnt.
Thc off-board proccssing may include vision. navigation, and path planning among othcr AT topics.
‘I’hc vchiclc should be vcry sirnplc to build, control, and maintain. It should bc flcxiblc cnough to usc indoors
or out, and capablc of carrying a varicty of cxpcrimcntal equipment. Calibratcd stcering and travcl arc
rcquircd for accurate movements. Considcring thc computation time necessary for navigation based on visual
scrvoing,’ high spced is unsuitable.
Mancuvcrahility is required to facilitate research in crowded environments such as those typical to indoor
cxpcrimcntation. Indoor USC provides the convcniences of close physical proximity of thc vehicle to the
rcscarchcrs, and the possibility for a controllcd environment. The additional capability for outdoor use allows
morc cxtcnsivc environments. These two abilities, togcther in a single vehiclc, allow thc investigation of a
broader sct of topics.
For sizc and maneuverability the vehiclc should:
1. fit through doorways,
2. turn around in hallways,
3. steer bctwcen obstacles, and
4. ncgotiate mildly rough terrain such as grass and sidewalks.
The width of a small closet door may be 24 inches (60cm), and a small hallway in an office building might be
as narrow as 48 inches (1.2m) wide. ‘I’hcrefore, a width under 24 inches and a ‘curb-to-curb’ turning lcngth
less than 48 inches is desired.
‘To get in and out of tight places, the vehicle should be able to turn sharply and reverse direction; ’large’
compliant tires are appropriate for bumps and uneven ground: and it should be strong enough to propel itself
and any attachcd equipment up an incline of say, 10 degrees.
Our schcdulc required designing and building a fully functional vehicle within a few months, a major dcsign
consideration. Our A I rcsearchers required a vehicle which could be easily controllcd and which would
perform consistently and without breakdowns. Therefore, the design had to employ tcchniqucs which
minimizcd both the time for fabrication and the propensity of the hardware to fail. The rcsulting design
shows clcgance in vcry simply and reliably fulfilling thc hnctional requirements.
’With our current Vision software, we can move about one-halfmeter at three minute intervals.
4
I
=III=
+
II=
I
r
n
Figure 2-I: Neplune
3. Design Decisinns
3.1 Cliassis and Suspeiision
‘1’0 simplify suspension, s tricyclc configuration is uscd. ‘I’CII iiich (25cin) diamctcr pncumatic tires2
;:IT uscd
to providc spring, complinncc, and traction on soft ground. Sincc the vchiclc will Uavcl at walking q ~ c and
d
slowcr, no additional springing or damping is considcrcd ncccssary.
Onc fork-mountcd whccl is uscd in thc front, and two parallcl whccls in thc rcar. To nccornmodatc sharp
L
at Icast 90’ lcft or riglit, and the whccl can
turning, thc stccrcd whecl is thc drivcn whccl. l‘hc fork C ~ I turn
bc drivcn forward or back, which cnablcs the vchiclc to rotatc about a point locatcd directly bctwccn thc two
rcar whccls. This is a prinic fcature in the dcsign.
A ’curb-to-curb’ turning lciith of undcr 48 inches is required. ‘I’hc whcclbase is 22.5 inchcs (57cm) and the
rear track is 19 inchcs (48cm). With thc 10 inch tircs, thc ovcrall width is 22.5 inches, and thc lcngth is 32.5
inchcs (83cm). With allovianccs for the fork, this gives a turning length ‘curb-to-curb’ of 42 inchcs (1.07111).
‘I’hc section on Structure provides more detail.
3.2 Tethcr
An untcthered vchiclc is not a requirement for much of the basic experimcntation. It is ccrtaiiily desirable to
havc wircless vehiclcs with the flexibility of path traveled and the extended range. Howcvcr, with this vehicle,
an umbilical is used in ordcr to minimize the number of subsystcms.
3.3 Power
‘I’o eliminate on-board power storage and recharging, powcr is supplied through the umbilical. In addition,
by using 12Ov~cmotors, the vcomplication and weight of power conversion equipment is climinated.
‘I’hcrefore, 12Ov~cis distributed for all on-board electrical equipment via outlets mounted in the vehicle
P
breaker protects against overloads. Each piecc of equipment provides its own power
frame. A ~ S A Mcircuit
convcrsion/protection. For more detail see the section on Electrical Control starting on page 19.
3.4 Motors
Synchronous motors were chosen for drive and steering as this allows elimination of feed-back and scrvoing
systems. ‘The motors are run for a length of time, and the revolutions are calculu/ed. This tcchnique, without
optical encoders or resolvers, enhances reliability.
One motor is used to propcl the vehicle. It is a 72RPM synchronous motor,3 that dcvelopcs 1800 o z h . of
torque, and weighs about 45 pounds (20kg). It is coupled to the drive whecl with a 3 : l reduction. The
~ ~a substantial
~
torque. It is
stccring motor is a gcarhcad synchronous motor4 with a final speed of 3 . 3 2 and
coupled to the steering neck with a 1:1ratio.
%ires are 10~3.40-5pneumatic hand truck tires rated for about 200 pounds load each.
3A Slo-Syn SS1800, the most powerful, commercially available, 1 l o V A C synchronous motor, with a short delivery time.
4Slo-Syn S s 4 ~ - P 2 , 4 0 oz.in.
0
at 7?RPM with a planclaaly reduction gear giving about 8000 o z h . at 3 . 3 2 ~ ~ .
6
For study, it inay hc itppropriaic to further rcducc tlic drivc spccd. u hilc inosasiiig thc stccring spccti, h i s
;illowing morc Iluid i~iovcs.‘I’licdrivc spccd inay hc rcduccd to 67.5% of its prcsciit 1 I+s b y sulisiitiiling thc
:Ipprol)riittc stock spi.ockcls. Adding an additional stngc ofstcp-down gcirriiig to LIIC drivc can rcclucc thc linal
speed s~ibslnnti;illy(also cffccting a substantial increase in torquc). As Ihc scnsory systcms bccoinc more
rcsponsivc, iin incrci\sc in spccd may bc prcfcrrcd.
3.5 Control
An on-board prtxcssor acccpts commands from an RS-232C link. ‘This prtxccssor controls switchcs and
rcports on opcration of vchiclc componcnts. The direction and duration of nin for cach motor is controllcd,
and thc fork position is monilorcd via limit switchcs. To sclcct onc of thc camcras a rclay is switclicd. ‘l’his
proccssor may also providc control and monitoring for other vchiclc-mounted cquipmcnt. For itiorc dctail of
thc on-board cornputcr SCC the scction starting on pagc 25. A mcans of manual control is also providcd.
3.6 Communication
Togcthcr with thc powcr, the umbilical carries a number of cables:
e a coax
for vidco signal from the cameras,
e a coax for Sync to the cameras, and
0
a multi-conductor cable with shielded bundles.
Serial communication is carried ovcr the multi-conductor cable. Spare conductorj allow for cxpansion. Once
thc commitment is madc to an umbilical, the overhead for communication to any device dcsircd to be
mountcd on the vchicle is low.
8
n
\
I
Figure 3-1 : Neptune - Cutaway
n
9
4. Structure
Ncpfune is madc from two basic asscmblics, thc fork and the frame. 130th parts wcrc dcsigncd to havc an
cxccss of structural fortitiidc to withstand abusc and providc secure mounting points for ituxiliary
cxpcrimcntal cqiiipmcnt. Aluniinum was chosen for structural componcnts duc to its mitchil1sbility. which
facilitated quick construction and allows subscqiicnt inodification to be niorc casily accommodatcd.
r.-
- -
i
i
i
i
t
ao
.
e
e
e
e
e
*
e
.
!
Figure 4-1: Nepfune- The basic vehicle (port side view).
4.1 The Fork
The drive motor is incorporated into the fork, dircctly above the drive wheel. This simplifies the transmission
of drivc power. Because this motor weighs almost 50 pounds (20kg) and devclops quite a bit of torque, the
fork was designed for strength: All thc aluminum used is 6063-TSor stronger. There are four pieccs: the top
and two side plates are .250" thick; the piece which forms the front and back is .125" thick.
The sidc plates arc square at the top and raper and curve at the bottom. They are welded to the top plate
inboard a few inches on each sidc. When rhc piece which forms the front and back is bent and wclded
around, an angle is formed which provides great stiffness to the whole asscmbly.
10
‘I’hc drivc motor is nlountcd to thc stiirhoilrd platc and can bc slid u p ;tud down to adjilsl thc lcilsion ol‘ the
drive chain. A sitnplc scrw jack is pl-ovidcd on this platc Lo aid adjus~iiicntof thc Iiwtor with ii wrc‘ncll.
‘I’hcrc is a reasonable amount of room hcrc to includc clutchcs, additional gcaring, ctc.
‘I’hc port plate has a cutout through which the back end of the drive niotor protrudcs. 111diis position, thc
motor and drivetrain arc well balanced abovc the drive whccl. Iklow this cutout is an ,idjustable bracket to
c
taking some stress off &IC starboard side.
help support ~ h motor,
One bearing flange block on cnch side platc supports the drive axlc. I’hcrc is room on thc port side, around
the drive axlc, for an accessory such as a rcsolvcr. Side covers provide protection.
The fork top is bolted to a steel flange which is welded to a hollow steel neck. ‘Iherc is no castor on the front
whccl. This was the decision because the vehicle drives slowly, and the geometry ol‘ thc stccring affects
calculation of direction, and thc drive motor is heavy. ‘The fork neck extends up beyond thc top frame piece.
A hole in thc sidc of the neck allows the cable from the motor to pass, up through the neck, and out abovc thc
top frame section. This cable simply winds and unwinds as the fork is turned. A round platc is scrcwcd onto
to top of thc fork neck for mounting of sensors or perhaps a manipulator. ‘Iliese wi!l turn as the fork turns.
4.2 The Frame
The frame is made of 4 inch square, extruded aluminum tubing (6061-T5),
with a .125” wall thickness. There
are four pieces. In addition to the high strength to weight ratio, the tubing provides a convenicnt raceway for
cabling and housing for small components.
The top section provides mounting for the stccring motor and the fork. As the fork is heavily loaded with the
drive motor, the steering motor was placed aft as much as possible. The stecring motor mounting provides for
stccring chain tension adjustment. The chain runs inside this section, thus providing protection. The fork
neck is supported with bearing flange blocks above and below. Snugly fitting end caps, fore and aft, provide
additional torsional stiffness to this member. These caps can also be considered as strong and adaptablc
chunks of metal on/in/from which to mount accessory equipment. For instance, the current camera support
is mounted to one of these.
The vertical section attaches to the mid-part of the top section. It has a lower end cut at 45’ to mate with the
next section towards the back.
The rear section is attached like a “T” and spans across the back. It has end plugs, again for torsional stiffiiess,
which also serve as strong mounting for the two pillow blocks which support the rear axle. These pillow
blocks are a variety which have rubber in them, and provide additional shock absorbing.
Holes
A number of 2.75 inch diameter access holes are let into the sections of the frame at convenient places. They
each have four capped holes around them to affix coverplates. Consideration to stresses in the frame dictated
placement. A line o f holes is provided on each side of the top frame piece to facilitate mounting auxiliary
equipment. Other holes are provided for passing cables. The frame, and the fork (see above) have been
designed to be robust enough to accommodate many additional holes, to facilitatc the mounting of auxiliary
cquipment.
c
11
(kunlerwc4ghts
N C ~ I U Ihas
W a fairly high ccntcr of gravity duc primarily to tlic position of Uic drivc motor. Considering thc
addition of othcr cquipmcnt, this ccntcr of gravity may bc madc highcr still. Whcn thc fork is stccrcd at right
anglcs to Ihc rcar whccls, thcrc is a possibility that thc vchiclc will fall ovcr bccausc of thc vchiclc’s narrow
track and (hc instanhncous starting naturc of thc synchronous drivc motor. To avoid this. countcrwcights arc
hung bclow tlic rcar axlc. ‘Ihcsc arc simplc rcctangular platcs of stccl. thc sainc s i x as thc rcar franic scction,
4 inchcs x IS inchcs (IOcni x 38cm). with mounting holcs aligncd with tllc rcar axlc pillow blwk holcs. Each
is 0.5 inchcs tliick (12.7nim) and wcighs ovcr cight pounds (3.7kg). ‘I’hrcc wcights pmvidc cnough
countcrbalanct! to prcvcnt thc basic vchicle from tipping, UI resf, whcn on a surfiicc of 45’. With this ballast
thc basic vchiclc. wcighs in at about 175 pounds (80kg). Additional compensating weights may bc hung hcre
whcn thc vchiclc is morc highly loaded.
i
i
I
Figure 4-2: Nepfune- Aft of port. (Counterweights not installed.)
4.3 Construction
The frame and the fork assemblies are welded for strength and simplicity of construction. A few thrcadcd
holcs were made in the fork side plates for jigging before welding.
Off-the-shclf housed bearings are used to mount the three rotating shafts (the two axles and the fork neck). In
12
'I'hc chain sprockcts are all kcycd to thcir shafts with stock split-tripcr bushings. Tlic bwring howings, fr:imc
caps, motors and thc rest of thc piirts arc bolted or scrcwcd togcthcr. Oncc all ihc picccs wcrc imclc
(including sizing tlic chains), assembly o f thc mcchanicd parts rcquircd lcss than c? miti-wcck.
Full shop drawings arc providcd in Appendix I.
4.4 Phase Shifting Coniponcnts
E;lch motor has two windings; onc rcccivcs thc AC current directly, thc othcr through a pF,m shifting rictwork
(a rcsistor and capacitor). Thcsc componcnts arc mountcd ncar their respective motors. Whilc tllc stccring
motor has componcnts small emu& to mount inside the vcrtical frame section, thc drivc motor hns a
capacitor mcasuring 2.5" x 5" x 7" (6cm x 13cm x 18cni). arid a 40Q rcsistor of 375 Watts!
4.5 Manual Control
A pendant control box is provided to allow for manual control (such as positioning the vcliicle in a convcnicnt
spot before starting an cxpcriment and parking it afterwards). Its functions include switches for turning the
steering motor lcft and right, powering thc drive motor forward and back, and emcrgency stop for safety.
Also includcd in tliis hand-held controllcr arc indicator lights which show the status of thc limit switches. A
connector is provided on thc vehicle to accept the controller, however. it may also be used at the othcr cnd of
the umbilical.
Manual control is also possible by scnding appropriate commands to the on-board processor via the RS232-C
link.
4.6 Liniit Switches
Thrce limit switches provide calibi-ation information for the fork. A Ccntcr switch, a Left/!!ight switch. and
an Out-of-Rangeswitch.
The Center switch has a dcadband of about +4'. This allows for:
1.the +lo play in the fork due to backlash in the steering planetary gear, and
2. overrun of the steering motor.
As thc backlash in the stecring gear incrcases with wear, a more accurate indication of whecl centering can be
had by monitoring transitions of the Lcft/Riglit switch whik in the ccnter range.
Out-of-l<angc is indicated when the fork turns more than 135' in eitlicr direction.
'Pfith the drive and steering chains are 40LL which is. 0.5 inch pitch and lifetime lubricatcd (to minimize rcquircd maintenance).
13
Front
1
2
3
0
0
0
n
i
o
0
1
1
0
1
0
1
0
1
1
1
0
1
1
1
0
1 O f f Limits
2 Center
3 RightILeft
legal states
UOT(
I AND 2 )
Illep<il states
S w i t c h e s 1 and 2 c a n n o t b e on a t t h e same t i m e
Figure 4-3: Limit Switch Ranges
4.7 Camera Mount
M o u n t i n g F l a n g e Removed
Figure 4 - 4 Neptune - Camera Support
A mounting system from a professional tripod manufacturer6 is employed to support the two tclcvision
cameras rcquircd by the current Vision research. The flange, couplings, lengths of tubing, and camera
mounting heads allow considerable adjustment of the cameras up and down as well as inboard and outboard.
This system provides great flexibility for mounting a wide range of cmcras and other sensors.
‘llavis and Sanford
14
Figure 4-5: Neprune - The first run
Kevin Dowling attending.
15
5. I I\;lcctianical
With i1 3: 1 chain reduction from thc motor. the drive ivhccl runs at 24KPh.1. This is about one foot pcr sccond;
less than 1 W l f . A set ot’sproclcts rcducing the ratio to 4: I is also available.
With the drivc motor providing lS00 ozin. of torquc and the 3:l reduction gcaring, about 67 pounds of
draw-bar pull is calculated. ‘I’his should takc a 350 pound vehicle up about an 11’ slope. A n carly test
involvcd loading Neprune to achicvc a coinbincd gross weight of about 350 pounds, starting on a lcvcl :;urfacc,
and driving it up a 9’ slope in a hallway near our lab. It did just succeed?
‘I’hc nature of a synchronous motor is such that it can drive a gtciltcr load than it can start. ‘I’hcrcforc, a more
uscful number is the actual slope on which the vehicle can be s/uried. Fully loaded with control box,
proccssor. cameras, ctc., Nzplunc weighs 175-200 pounds. For tcsting, a variably inclincd surfncc was uscd.
Experimentally, it managed to s m l on a 10’ slope.
Another ‘feature’ of these motors is that once they slip, givcn enough spccd, they will turn relatively easily.
Whilc this may be a convenicnce when turning the fork by hand (without power), it may be hazardous to
stand c?clowthe vchicle when attempting too great a slope. Caution is recommended.
‘The steering motor with its 3.32RPM output and a 1:1 chain drive to the fork neck, provides more than enough
torque, even when the vehicle has no forward rolling speed. Final steering is at about 20’ per second.
5.2 Moving Accuracy
Synchronous motors have a starting delay which varies somewhat. This is not apparent in the steering which,
due to the low gcaring, has a measured accuracy within one degree. On the other hand, effects of this delay
arc significant in thc movement of thc drive motor.
Measurements were taken for straight moves forward and backward at varying lengths of time. An equation
for calculating distance travcled is
distance = s p e d x (time - delay)
where speed is 12.3 inches per second and deluy is considered 0.12 seconds. Variation in short moves (one
secund) was about k0.75 inches, and in longer moves about k0.5 inches of the desired position. Different
algorithms for moving will result in different dead-reckoning accuracies. One early method for moves
required very little geometric calculation:
1. Turn the steered wheel 99’.
2. Run the drive motor for a short time to rotate the vehicle about the center point of the rear axlc.
3. Return the steered whecl to 0’.
4. drive forward along the diagonal formed by the desired change in the x and y positions.
5. Again turn the stccrcd wheel 90’.
7110wevcr,this test was not optimal as it was a live load, namely Hans Moravec
16
0. 12!)tatc tlic uchiclc to rJic clircction it originally started.
7. Again str-;iightcn llic stccrcd whccl.
If at rhc sccond step. l l i o intcnt is LO rolatc thc vchiclc n only small amouni, thc strirtiitg dclay vari,ition o f d ~ e
driving motor will inorc adwrscly afl'cct movc accuracy.
For cach niove cmploying this rncthod. the stcercd wheel is scnibbcd back and f011h a numbcr oftiiiics by thc
substantial cncouragcmcnt of thc stccring motor.8
Anothcr iilgorilhrn cinploys S-shnpcd curvcs (fitting splines to positions along Lhc dcsircd path) which ;dlows
thc drivc motor to run continuously while thc steering motor is moved back and forth. 'I'his hclps to
minimizc thc cl'fect of thc drivc motor starting dclay crror.
8Anothcr useful application of this ability (to rotate the front whecl in place) is to orient a scnsor or manipulator atlaclied to the fork
asscmbly.
17
5.3 System I’crformance
Thc physical vchiclc bccomcs a robot with thc
addition of cnvironmcnt mapping ciipabilities
[7,4], path planning [6], and navigation [l],
software. Ihe currcnt wftwarc runs off-board on
a VAX-11/780. In a numbcr of Vision bascd
expcrimcnts (using a binocular pair of
monochrome vidco camcras), thc vchiclc made
moves of about 20 inchcs (0.5m) at thrcc minutc
intcrvals, to successfully navigatc 32 fcct (1Om)
through a room of obstaclcs. Fxperimcnts have
also been run with cnvironment maps dcrived
from a 360’ ring of 24 sonar range sensors
mounted on-board.
Near-fbture plans include an on-board video
digitizer and processor for some of thcse vision
functions.
5.4 Summary
To facilitate autonomous mobile robot research, it
was decided to build an uncomplicated mobile
basc. Ninety days later. a tested, working vehicle
was used for vision and na-Jigation experiments.
Its operational simplicity allows it to be used for a
variety of different studies by a number of
Its mechanical fortitude
researchers.
accommodates the mounting needs of many
different experimental apparatuses. By w e h l
design and employing standard components
wherever possible, this rugged, reliable vehicle
was produced in a minimum of time.
Figure 51: Nepiune - Navigating in the Lab
6.11Slectsical Control
Simplicity and thc ilbility to changc hardwarc wcrc primary dcsign considcralion:; in tlic clcctrical system.
All control is bin;ii*y. I .ogic signals from thc on-koilrd prtxcssor control solid-sratc relays which s\vitcli powcr
to thc motors. Somc additional circuitry providcs for safcty:
'I'hc systcm must bc powcrcd ON and thcn thc S'rAR'r button prcsscd bcforc thc motoIs can bc
cncrgizcd.
0 130th windings of onc motor cannot bc switchcd on siinultaneously.
0 A dclay prcvcnts a motor from bcing instantly rcvcrsed.
0 A numbcr of ~'1'0~
buttons hclps prcvcnt collisions in an cmcrgency.
0 Thc drivc motor is used as a brakc whcn the vehicle is not moving by applying a IX 'holding'
voltagc.
0
6.1 System Power
A latching configuration is uscd for turning on system power.
11OVAC
c
Hot
Motors
DC H o l d i n g e t c .
5v
AC
C a l e x 1.5.500
l e ut r a 1
-E
+
-
-
Cmn .
-AC
I
%
r
DC 200P
C h a s s i s Ground
controller
Start
Latching C i r c u i t r y
I
__Q_LI>_
I
I
stop
stop
I
Figure 6-1: Start/Stop System
The start/stop system must have the following properties: When a stop switch is prcsscd, power to the
controllcr and motors is shut off immediately. If the start switch and a stop arc pressed sirnultancously, thc
powcr stays off. The manual controller must be able to disconnect froin the vehicle and yct, when connected,
bc able to stop the vehiclc with the stop switch provided.
'I'hc figurc abovc shows a simple outline of how this was accomplished. The inputs to thc latching circuitry
arc thc start and scvcral stop switches. When the start switch is presscd, thc output linc is pullcd low, thus
turning thc control side of the two SSR's ON. Onc SSR turns on thc AC power for thc rcst of thc controllcr
and the motors. Thc othcrs turns on thc 5VDC for thc logic in the rest of thc controllcr.
20
'I'lic l:i!.cliing citxuicry c;in bc sccn in dctaii in Appendix 11. A sirnplc flip flop w a s constructcd from NOR
giiics. I n addition, thc sucitch inputs arc gatcd in sucli ;I wily [tiat illcg;tl inpiits to this conligiiratioti cannot
(xcitr. 'l'lw oiqm latches high if any of thc stop switchcs arc prcsscci. thus t.urning off thc coittrol side of thc
SSl?'s. If tlic inanual controllcr is conncctcd, and thc stop switch on it is prcsscd. it pulls a n input low. 'I'hc
arnngcmcnt of discrctc cornponcnts shown with Schmidt trigscr buffers is IO initi:ili/.c tlic statc of thc circriit
when powcr is itpp!icd to Ncptunc and make 3urc thc SSR's rcmain off until thc siart switch is prcsscd.
6.2 R.lolor switching
Sevc'raI solid-statc relays provide a sirnple intcrfacc bctwccn thc logic lcvcl signals and switching AC powcr to
thc motors o n and off. l'hc rcqiiircrncnt was a 10 amp capacity with liighcr peak ratings and a voltagc rating
able to handlc thc doubled voltagc that can occur across thc phase shifting capacitor and the motor windings.
7'hc rnodcl Opto-22 2401)lO is uscd for Oircctly switching thc motor currcnts for both motors,
Scc Appendix I1 for the schematic layout of the SSRs and tlic motor windings.
Whcn thc drivc motor is off and no currcnt is flowing though the windings. it bccotncs fairly easy to turn.
This is unacccptablc sincc rcliancc on position is csscntial and tlic vchiclc will bc stopped uii inclincs. On the
othcr hand. thc stccring motor is gcarcd down about 20:l so that a great deal of external forcc is rcquircd to
rotate thc fork when it is not powered.
For tlic drive motor, a DC current is put through the windings which holds the motors with a substantial
torquc. The motor manufacturer recommcndcd a 9.5VDC supply capable of handling 4.7 Amps. A ~ V D C
unrcgidiltcd supply, ratcd at 10 Amps, is used bocausc exact voltagcs and rcgulation arc not neccssary for the
holding torquc. In a simple experiment, Nepfurie was placcd on a stccp 30 dcgrce slopc, and thc I>C currcnt
put across onc of the motor windings. Nepfuriewould tip over before tlic whecl evcr gave way.
When switching motor dircction by turning off onc SSR and turning on another, carc must be taken during
this transition period because the capacitor may discharge through the two SSRs. This short circuit discharge
currcnt may destroy thc SSRs. An additional resistor may be placed in series with each SSR for protection,
but if the control circuit insures that one SSK is off before the other is on the resistors are unncccssary. A
sirnplc dclay circuit has been incorporatcd into the control signal circuit to do exactly this. It is possible to do
this in software by timing the output control signals but the potential is there for damage. Thus, thc hardware
incorporation of the dclay is a major safety factor.
6.3 Digital Control Logic
All logic in the Neptune Electrical Control System operates on active low logic. That is, things turn on and
switch with logic 0 signals. This is consistent throughout the design. The simple reason is that thc digital
circuit clcmcnts cffectively sink more current than they can source.
'Thc inputs to the system are the processor signals and the manual controller signals. 'The outputs are the
motor signals and directions. In between, there is hardware to make sure that conflicting signals aren't
gcncratcd (c.g., a motor told to go forward and backward at the same time). and that manual and computcr
control do not conflict.
21
6.4 1)cl;iy Circuit
‘132
i r
Time delay
Figure 6-2: Delay Circuit
The delay circuit provides a safety factor when motor direction is changed, ensuring that 0111: SSR is off
before tl;c other is turncd on. (See the scctioii on Motor Switching for morc iiifurinatiori on chis.) Ihc dc!ay
circuit has thc following property: When an input signal drops Jow (active), thc ouipiit signal will drop low
slightly later. When the input signal goes high, the output signal also goes high (cffcctively at rhc same tinic).
This circuit block buffers and inverts the input signal. The inverter output fans out to a rcsistor-capacitor
arrangemcnt and to thc input of a Schmitt trigger NAND gate. The XC arrangymcnt delays thc signal because
the capacitor charges up over a finite time, and the resistor impedes the current to it. l’hcn after soiiie time,
the capacitor charges up and the NAND input on the RC circuit side changes state. ‘I’hc Schniitt trigger
output changes smoothly due to the hysteresis effect. When one illput is low. a property of NAND logic
causes t’lc output to go high. Thus when the delay input goes high, the inverter output gocs low, thcn oile of
the NAND inputs is low and it’s output goes high. ‘I’his simply means that when the input gocs back to a high
state the output also returns to a high state.
l’hc values of R and C are common values that provide an RC value of 0.08 seconds (assuming actual values
are equal to their labeled values). Originally the delay was longer with an ovcrestimated safety factor but for
smooth motions while turning the delay was reduced to the present value.
6.5 Mairi Control Circuitry
As dcscribcd, each motor SSR control has a dchy unit associated with it. ‘l’he delay unit hac computer or
manual controllcr signals as its input. A simple methcd of switching between thc manila1 controller and the
computer so conllicting signals will not occur is obtained by using a Quad Dual Input Multiplcxcr. A 74C157
is uscd and fits the bill perfectly. ‘the four inputs are: Drive Forward, Drive Back, Steer I,cft, Stecr Right for
each set of inputs.
22
N(;tc ( h i tlic n i , i i d controllcr has a switch to cnablc m:i~iuA c o ~ ~ tI! ~~ I~. I JlI S. thc SCICZ~ ('tl thc I I I ~ I XC i i i l b l c
1t)w 111us ,~llowiiigthc sct ol' four inputs froin the mnnual controllcr to control thc vc4iicle. Whcn lhis switch
cwablc high tlic coinputcr signals control llic vchiclc. I%ssivc ~.,ull-upresistors arc used for
Ixiiigiiis tiic signal high whcn somcthing is not cnablcd.
pills thc
iitiix
'I'hc coinputcr inputs still havc thc capability of turning on both SSRs to n motor io wftwarc. 'I'lirougli thc
xitiition of a I'cw K A N D gatcs, this both-on skitc is not allowcd to occur. Wlicn it. docs, thc motor is turncti
off bcciiusc both $pals arc low.
l)uc to thc physical construction of thc switch. thc manual controller cannot ha4c both directions iiiriicd oil at
oncc. ' I ' t i c final addition is NANDing thc inpuls to thc drive motor to scc when thc L
X Holding 'I'orquc
should bc turncd on. If both dircctions arc turncd off, lhcn the IIC SSR should bc tuincd on. Onc NAPID
gatc suMccs for this.
rclay. Thc relay has logic levcl voltagc input
I h c IX holding torquc is switched by a small clectrc~-mecIiai~ical
to a small 2N2222 transistor configuration which drivcs the rclay coil. An ilttempt to find a IX SSK with the
com1iin:ition of lVl'L level control, high current handling (SAMP), and high blocking voltage (200~)was
unsucccssfi~lso a simple rclay configuration was uscd (SCCAppendix 11).
Camera Switching
Coax R e l a y
status l i g h t s
I
V i d e o Out
computer
-
Manual
Camera
-
5 VDC
Figure 6-3: Camera Circuitry
'I'wo cameras arc mounted above the vehicle for capturing stereo image data. Instead of running a coaxial
cablc for each camcra, one cablc and a mcthod of switching betNccn the two cameras was designcd. A
processor signal turns on the input sidc to a DC SSR. Tliis SSR triggers a coax relay which switchcs the output
line bctwecn camera 0 or camcra 1. The rclay is a doublc-pole, doublc throw design. The othct pole is used
simply to sclect ari LED output to signal visually which camera is currently being selcctcd. ?'he LEIYs are
mountcd adjacent to the DNC conncctors to which cameras are connected to on the rcar of thc control box.
In addition, a pair of switchcs enables manual camera switching. This makes it easy to do calibrations, inspcct
23
cancr:i fi91s
CIC.
0nc switch swilcixs from coniputcr control to ni~inu:rlcontrol Lind thc ollicr sclccts onc of
Ihc t w o C ; I I I I C ~ ; R Ag,iin, thc l,l:lYs iiidicatc which camera is currcntly sclcctcd.
6.6 Layout
A11 the powcr supplies, circuitry. switchcs, ctc. fit into ;I 1 2 ~ 1 2 x 4inch box that fits on tlic back o f N ~ p t i u i e .
SCCthc drawings in Appendix 11 for further layout dctnils. Considcralions for c~icIosurcs,switchcs, and
conncct ors wcrc si in pl ici ty and rugged ncss.
6.7 Possible Enliancerncnts
Bccausc thc drivc motor is synchronous, Nepfune starts with a slight luirh whcn tlic motor is cnablcd.
Itamping up thc frcqucncy is a possibility. A simplc circuit using powcr MOSFE'l's m a y bc used to chop R 13c
input voltagc and provide logic lcvcl control of speed. There are on thc market variable frcqucncy drives but
with fcw cxccptions, thcsc arc dcsigncd for multiplc phase induction motors and arc physically fairly hrgc for
Ncplunc to haul around.
In scrics with thc stop switches is planned a bumper "skirt" to shut down the motors whcn a collision is
imminent.
24
25
On board, Nqmrir carrics one micrt~omputcr spstcm, built around a Motoroln MCOY000 32-bit
rnicroproccssor. ‘I’hc liinction of tlic microconiputcr is to act as an intclligcnt intcrfxc t)ctwcen the off 1)oard
computers (Le., ii VAX running vision softwarc) and thc on board hardware, chicfly the niotor drivcr and thc
canicra sclcction circuitry.
T w o RS-232 lincs in tlic umbilical arc uscd to connect thc niicrocornputcr scrially to thc outsidc world. Chic of
tlicsc lincs connccts thc ~~iicrwomputcr
to a debugging and stiltus tcrminal, whilc thc othcr connccts it to thc
\’AX. ‘I’hc microcornputcr is conncctcd to thc various robot control and status lincs through a parallcl port.
Figurc 7-1 dcpicts thc basic micrtxomputcr’s architccture, along with the extcrnal conncctions.
Microprocessor
D r i v e Forward
D r i v e Backward
Steer L e f t
Steer Right
Camera S e l e c t
Right Switch
Center Switch
Out-Bounds S w i t c h
I
To VAX
Figurc 7-1: On-Board Processor Block Diagram
...,
To allow the VAX to control the robot, the microcomputer runs a special controller program, ‘Ncpcon’, H h
parscs high lcvcl commands from the VAX and translaics them in to robot motion and control. Appcndix I11
contains a completc description of the ‘Nepcon’ program.
Physically, tlic microcomputer is built on one Augat Multibus prototyping board, approximately 7 by 12
inchcs. To minimize power consumption and possible heat problem, CMOS integratcd circuitry was used
whcrc cvcr possible (the only non-CMOS circuits arc thc 68000 microprocessor, thc 68230 parallcl
intcrfxc/timer and the RS-232 bus drivers). ‘ h e microprocessor is enclosed in a 10 by 14 by 4 inch
aluniiiiuin box, along with a small Calex 5, 12 and -12 volt supply.
+ +
26
27
8. Bibliography
Mntthics, I,. CYC ‘I’horpc,C.
b”I10 and ‘I’YCIIO:Vision and Navigation for Mobilc Robots.
In I’rocwdiiigs r~11~/.:I:‘-Occari.~-H4.
I EEH, 1984.
Moravcc, Hans P.
Obstuc-le Avoidaiice and Navi,qcilioti in (he Real World by a Seeirig Robot Rovrr.
Phi> thcsis. Stanford 1Jnivcrsily. Scptcmber, 1980.
l’ublishcd as /<obut Rover Visual Navigalion by LJMI Kcscarch l’rcss, Ann Arbor. Michigan, 1981.
Moravcc, Hans P.
Three Ihgrees for a Mobile Robot.
‘I’cchnical Ikport, Carncgic-McllonUniversity, May, 1984.
Moravcc, H. 1’. & Elfcs, A.
High I<csoliitionMaps from Widc Angle Sonar.
I n Proceedings of IL:‘EE-Itilernalional Corference on Robotics and Arrlomutioti. II-XE, 1985.
Muir, P. F.
Dcsign of a Digital Scrvo Controller for Brushless D.C. Motors for Motion Control ol‘a Mobilc Robot.
Mastcr’s thcsis, Carnegic-Mcllon University, December, 1983.
Thorpc, C.
Path Kclaxation, Path Planning for a Mobile Robot.
In Proceedings of AAAl-b4. AAAI. 1984.
Ihorpc, c.
An Analysis of Interest Opcrators for FIDO.
In Proceedings of IEEE Workshop on Computer Vision. IEEE, April, 1984.
Appendix I
Appendix I
Mechanical Drawings
Appendix I
t
d
4
N
Y)
10
c"-
I2*O
n
-I I
7
.250
.126
-/
.250
.II
IL.
*
-
nn
uu
P
0
6.00
2 . 1 2 5 1
2.125-
"f
of
O-%
-
2
?
m
0
v///
1
/ A I//////
L
I
Four h o l e s
I n a square
.326 d l a .
0
One h o l e
Centered
.600 D l a .
Chamfer edges.
Y i/ / A
I/
I,
I
f o r IYB s c r e w s
/i/i/
/ / / / / I
i
///I
I//
i
N o t e : A l l measurements: + / - . 0 1 5 "
Or w h e r e g l v e n .
N o t e : Make o n e .
Materlal:
. 2 5 0 " Alumlnum
6063-T6
0
Mobile Robot Lab
Tjt!e:
Front
'
4 . p
V
1984
T r i k e F o r k Top P l a t e
F i l e : [ r o v e r ] / u s r / g w p / d . tri I F o r k T o p . dp
Date:
7-Feb-84
Drawn by:
Checked b y :
I Scale:
G.Podnar
NTS A p p r o v e d :
Dwg,
No.:
Fork
&?
0
p--3 . 3 7 5
m
N
b
8.500
-
-
=
3.375
1
0
T-
Seven H o l e s
About 1 . 0 0 0 deep
D r i l l e d and
Tapped f o r #8
0
0
0
rn
I
EI
I
.......
- ....>
I
10
c:::
--
lb
0
-
i
7
Front
N o t e : A l l measurements:
Or w h e r e g l v e n .
+/-.OM"
1 Inch
N o t e : Make o n e .
N o t e : B e v e l p e r l m e t e r on one s l d e
f o r we1 d i n g .
Material:
. 2 5 0 " Aluminum
6063-T6
U
b
4 0 -
Q
0
N
hrl
Seven H o l e s
About 1.000 deep
D r i l l e d and
Tapped f o r #8
e
fs::
.......
Y-!
N
(
O
0
2
+
+ - I -
0
m
7
;
Front
N o t e : A l l measurements: + / - . 0 1 6 "
O r where g i v e n .
N o t e : Make one.
N o t e : B e v e l p e r i m e t e r on one s i d e
f o r welding.
The t a p e r i s d e t e r m i n e d b y a l l n e f r o m 8.500 t o a b o u t 14.6 I n .
w h i c h I s t a n g e n t t o tJe 3.500R a r c ( a b o u t .76 I n ).
The a r c i s a b o u t 156
.
Material:
. 2 6 0 " Aluminum
6083-TS
Mobile Robot Lab
1 inch
U
A
V
0
1984
T i t l e : T r i k e Fork L e f t Side
-
12.000
.
______(
2.62
fj-r
0
2 RI
Four holeseach end.
.ZOO d i a .
4
N
W
a
I
1
Four h o l e s
each end.
.I75 d i a .
Countersunk
one s i d e .
E i g h t i n - b o a r d countersunk holes are t o
mate w i t h t h r e a d e d h o l e s i n t h e
L e f t and R i g h i s i d e p l a t e s .
The p i e c e s w i l l be screwed t o g e t h e r
before welding t o help a neat job.
J
-
Four h o l e s
each end.
.175 d i a .
Countersunk
one s i d e .
-
3
r(
Io
3
One h o l e
each end.
.500 d i a .
Beveled
both sides.
Bend l i n e
Gradual b
s t a r t s her
T h i s p i e c e f o r m s t h e f r o n t and back
of the fork.
It w i l l be w e l d e d w i t h
t h e L e f t and R i g h t s i d e p l a t e s and t h e
Top p l a t e . The f i t o f t h e bends I s
important f o r strength.
I f necessary
t h i s p i e c e may be made i n two p a r t s
w i t h a s p l i t halfway (see * ) .
1
4
1
e
E i g h t out-board countersunk holes are
f o r brackets t o hold side cover p l a t e s .
-,
6.000
-
9
I
+
Front
N o t e : A l l measurements: + / - . 0 1 5 "
O r where g l v e n .
Note: Make one.
1 Inch
U
0
0
0
0
0
Material:
. 1 2 5 " Aluminum
6063-Tl
0
Mobile Robot Lab
T l t l e : T r i k e F o r k F r o n t and Back
Drawn by:
G.Podnar
b
4
-
f
0
1984
Top
J
Rignt
Front
Side
Front
(cutaway)
F i l l e t welds a l l t h e way aro'rnd t h e s i d e
p l a t e s on t h e Jut.-hoard s i d e .
Fillet w<!lds a l c n ~t h a i n s i d e c n ? n e r s w::era
t h e t o p piQ:;e n I C e t S t h e f r o n t dnd b a c k .
KO:C
t h a t t h e c e i i t r a l p a r t . g f t h e s e welds
drt! a c ~ ~ i i l ~ i i r ; h bt y
? dr e a c h i n g i n s i d e froni
t h e I'oiinded e n d .
The p i c c c s v i - i l I h e s c r e w e d i . o g e t h e r
b e f o r e w c l d i t i ! , t o h e l p il i l e a l j o b .
Claterial:
".*""
Cover B r a c k e t s
I
7
T
Two h o l e s
.180" d l a .
0
UJ
h
4_
0
t
W
r
m
UJ
n
rn
0
c1
In
0
t
0
0
1
I
N
W
d
I
h
M a t e r i a l : Alumlnum
Note: Make 10
I
one I n c h
I
0
In
J
Right
Note: A l l measurements:
O r where glven.
Front
1 Inch
+/-.015"
U
Note: Make two.
The t a p e r I s d e t e r m l n e d b y a l i n e f r o m 8.450 t o a b o u t 14.46 i n .
i n ).
w h i c h I s t a c g e n t :a tJe 3.450R a r c (about..75
The a r c 1 s a b o u t 165
Materlal:
b
.0625" Alumlnum
4
.
-
V
0
Mobile Robot Lab
19~4
T f t l e : T r i k e F o r k S i d e Cover
<-
F l l e : [rover]/usr/gwp/d.trl/ForkCov.dp
I
D a t e : 18-Feb-84
S c a l e : NTS Approved:
Drawn b y : G.Podnar
Dwg. For
No.:
Cheched b y ;
01
0
-0
0
f
0
0
.
n
*
0
Four Holes
i n four
corners.
(0
0
0
- 0f oped
r #E.
0
Y)
-
,Ld
1.365
0
0
tb
Two s e t s o f
t w o h o l e s each.
D r i l l e d and
tapped f o r 14.
-0
I
I
I
I
I
N
v)
0
I
-0
0
0
'9
I
I
3.375
0
ID
I
O
~
I
i
7.250
I
0
J--0
0
0
I
-0
I
I
I
I
I
I
-.
-*=
750
0
0
0
1.150
-0
T0-0
I
0
2
Two h o l e s .
D r i l l e d and
tapped f o r #4.
N
N
0
c 4
0
+
-0
2
Two s e t s o f 16 h o l e s each.
D r i l l e d and t a p p e d f o r U4.
One c u t o u t
r-.600
a
H
0
0
0
_.
Mobile Robot Lab
N o t e : A l l measurements: +/-.010"
O r where g i v e n .
Note: Make ons.
Material:
b
. 1 2 5 " Aluminum
€061
I
1 inch
I
f
-
F i l e : [rover]/usr/gwp/d.tri/ForkRes.dp
D a t e : 20-Feb-84
II S c a l e : U T S I ADuroved:
Drawn by: G.Podnar
Dwg* F o r k
No. :
Checked by:
..
0
1984
r-
1.375
% Hex Cap Screw.
I
2" l o n g . f i n e t h r e a d
With a reduced diameter o f
.U
.1840
f o r a l e n g t h o f ..ZOO
A l l edges
filed to
remove
s h a r p edges.
3.250
I
n
1
I
-k
I
2.750
J.mi
.1840
.625
One h o l e JAR. dia.
.100
.200 c l e a r d e p t h
0___/1
.
,
*
I
.
@
m
One h o l e
D r i l l e d and t a p p e d
f o r d f i n e thread
16
.
,
,
I
,
; ;
I
,
I
,
% Hex n u t
Two h o l e s
.426" d i a .
-
I
I
Assembled
f o r locking
Materials:
One 5/16" s t e e l hex cap s c r e w , f u l l y t h r e a d e d .
One 5/16"
s t e e l hex n u t .
Other p a r t s
-
mild steel.
Note: Make one s e t .
Ial
I
Title:
b
Mobile Robot Lab
T r i k e Motor Support A
For A d j u s t i n g Chain lension
File:
.500
Smooth p a r a l l e l s u r f a c e s
[rover]/usr/gwp/d.tri/ForkSupA.dp
D a t e : 17-Feb-84
I S c a l e : N T S l Approved:
Drawn by: G.Podnar
Dwg. F o r k
NO.:
nf
Checked bv:
1
t
T
3.260
A l l edges
filed to
remove
sharp edges.
One h o l e
1.0" d l a .
One h o l e
D r i l l e d and tapped
f o r 318" c o a r s e t h r e a d
U
.260
Materlals: Mild steol.
N o t e : Measurements + / - . 0 6 0
o r as noted.
N o t e : Make o n e .
S u p p l y one 3/8" x 1" Hex Cap Screw a n d washer
0
Mobile Robot Lab
TitIe:
1384
T r i k e Motor Support B
For S u o o o r t i n n R ? c k F n d o f M o t o r
..
F i l e : [rover]/usr/gwp/d.tri/ForkSupR.dp
Date: 17-Feb-84
I S c a l e : N;SI Approved:
Checked b v
I
NO.:
Fork
,
,
f
1
Hole bored t h r o u g h s h a f t
f o r passing w i r e s .
.
. 5 0 " d iameter
Concent. + / - , 0 5 0
See D e t a i l f o r
h o l e p a t t e r n and
size.
To p r e v e n t a b r a s l o n .
f i l e o r bevel sharp
edges a t e a c h end.
4
/- - 15 ~ ~ 4 5Chamfer
'
.500
I
I
I
I
p;
I
f
Hole bored
. 6 0 " deep
I n slde o f
shaft f o r
passlng
wires.
.50" d l a .
2
d
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
0
0
d
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
L
SICE ( c u t a w a y )
&x45'
4
.500
-
1
I
I
I
I
I
I
I
I
I
I
I
I
I
L
/
Chamfer.
HOLF PATTERN DETAIL
T h r e e h o l e s l o c a t e d 120 ' a p a r t .
D r i l l e d t o d e p t h o f about 0.6"
and t a p p e d f o r #8 s c r e w s .
r
K e v s e a t .260 x.126
Length: 3"mln.
4"max.
C e n t e r e r d 4 " f r o m end as shown.
K e y s e a t ends need n o t b e n e a t .
-
I
I
-
1
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-
1
I
I
I
Note
,I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-
I
I
I
I
Y Y
>-4
To p r e v e n t
abrasion,
f i l e o r beve
s h a r p edges
i n s l d e & out
0
I
I
I
1
I
I
I
I
I
I
I
I
I
I
I
I
I
BACK
I: Groove f o r r e t a i n i n ;
ANSI s t a n d a r d f o r 1
ring.
shaft.
N o t e 2: A l l measurements: +/-.010"
O r where g l v e n .
N o t e 3 : Have S h a f t made b y
Thompson I n d u s t r i e s , I n c .
Manhasset. NY 11030
(615)883-8000
N o t e 4: Make One.
M a t e r i a l : 60 Case h a r d e n e d s t e e l .
1" s h a f t l n a
Class "L** a i a . t o l e r a n c e .
Mobile Robot Lab
T i t l e : T r l k e F o r k Neck S h a f t
0
1984
As semb 1 y
Use F o r k S h a f t .
Square keyway w i t h f l a n g e h o l e s .
Four h o l e s
i n a square
.326 d l a .
One h o l e
Centered
.880
Weld o r b r a z e
both sides.
Bevel welds
a l l around.
A f i l l e t around t h e f o p o f
a t l e a s t .126 x . 1 2 6 .
dia.
.E76
Weld s t r e n g t h :
30.000 l b s . / i n ?
mln
Bottom f i n l s h e d by g r l n d l n y f l a t .
\ _ -x46
I
.I6
Chamfer.
One s l d e .
Note: A l :
measurements: + / - . 0 1 6 "
Or where g l v e n .
Note: Make one.
Materlal:
I
1 Tit;e:
1
4 . p
Q
.250" M l l d S t e e l
Mobile Robot Lab
T r i k e Fork Flange
F I1e : [ r o v e
1.1/ u s r / ywp /d
I
. t r i/ F or k F 1 ange .d p
Date: 18-Feb-84
S c a l e : NTS Approved:
Drawn b y : G.Podnar
Dwg. F o r k
Checked by:
No.:
.
.
-
f
3.825
- One Hole,
\
L
I
T h r e e h o l e s l o c a t e d 120 O a p a r t .
D r i l l e d and countersunk f o r #8.
N o t e : A l l measurements: +/-.010"
O r where given.
N o t e : Make Two.
M a t e r i a l : Alumlnum . 1 2 6 " 6061
Mobile Robot Lab
T i t l e : T r i k e F o r k Neck S h a f t CaD P l a t e
4
I
inch
'I
b
I
F i l e : [ r o v e r ] l u s r / g w p / d . t r i / F o r k C a p . dp
D a t e : 30-Jan-84
[Scale: NTS Approved:
Drawn by:
Checked by:
G.Pidnar
'
.
Dwg
No.:
0
1904
.625"
/-
UNC t h r e a d
L
N o t e 1: Have S h a f t made by
Thompson I n d u s t r l e s . I n c .
Manhasset. NY 11030
(616)883-8000
-I--
+
'9'9
b
--
.625" UNC t h r e a d
/-
N o t e 2: Wake One.
b
M a t e r l a l : 60 Case hardened s t e e l .
.626" !hafting
C l a s s L" d l a . t o l e r a n c e .
-I
4
I
I
1 Inch
Mobile Robot Lab
T i t l e : T r i k e Rear Axel S
sh
h aa ff tt
F i l e : [rover]/usr/gwp/d. t r i / R e a r A x e l . d p
D a t e : 1-Feb-84
S c a l e : NTS Approved:
Drawn by: G.Podnar
Dwg
I
'
Checked by:
-
.
No.:
0
198,
Chamfer b o t h ends
t o ease p r e s s i n g i n t o wheel.
-
-Allen
S e t Screw h o l e
D r i l l e d and t a p p e d f o r 1 1 0
A l l e n Cap Screw h o l e
D r i l l e d and t a p p e d f o r Y10
( a t same l o c a t l o n on o t h e r s i d e )
K e y s e a t : . 1 2 5 x . 0660
The l e n g t h o f t h e h o l e .
N o t e 1: Measurements +/-.010"
N o t e 2: Make One.
o r as
ndicated.
.
Material: Wlld steel.
0
Mobile Robot Lab
1984
T i t l e : T r i k e D r i v e Wheel Hub
F11 e: [ r o v e r ] / u s r / g w p / d
I
. t r i/Wlieul Hub. dp
D a t e : 2-Feb-84
S c a l e : NTS Appruved:
Drawn by: G.Podnar
Dug.
'
Checked by:
No.:
.-
+x45
0
Chamfer
0
901
ff:
N o t e 1: Have S h a f t made b y
Thompson I n d u s t r i e s , I n c .
M a n h a s s e t . NY 11030
(616)883-8000
N o t e 2 : Make One.
M a t e r i a l : 60 Case h a r d e n e d s t e e l .
0.6" s b a f t l n g
C l a s s L" d i a . t o l e r a n c e .
K e y s e a t : .128 x . 0 6 2 6
Ends need n o t b e n e a t .
b
V
-I
Mobile Robot Lab
T i t l e : T r i k e Fork Axel S h a f t
.600
I
F i l e : [rover]/usr/gwp/d.tri/ForkAxel.dp
%x45
Chamfer.
Drawn b y :
Checked by:
G.Podnar
No. :
1
1 inch
0
1984
I
\
Measurements f o r b e v e l s
a r e g i v e n t o edge o f
opening (bottom).
B e v e l i s made a: 4 5 d e g r e e s .
(May b e f i l e d . )
p-
7.750
3.875
_____de__
3.876
-
1-
-
I
I
' Bottom
\
'1
2
?
ID
L
M a t e r l a l : Aluminum
N o t e : Rake o n e .
Rs;I
0
Title:
File:
Date:
A" t h i c k .
16
u
Mobile Robot Lab
T r i k e Fork L i m i t S w i t c h C o n t r o l P l a t e
rr9verl/usr/awD/d.tri/ForkLiinit.dp
10-Feb-84
Drawn b y :
Checked b y :
IScale:
Approved:
NTS
G.Podnar
'
Dwg.
No.:
Fork
Of
I
7
u
4.00"
0
0
0
0
c(
-0
0
0
0
=-l-=-l"
-
0
o
0-
0Q
0
Six holes.
D r i l l e d and t a p p e d
f o r #6.
9
c)
-0
0
Four h o l e
0
M a t e r i a l : Aluminum 6 0 6 1
.le75 ( o r . 2 6 0 ) t h i c k .
N o t e : Make one.
.15Ox. 400
Two H o l e s
.150~.650
If 14 I
\ I
.3751
:1 I
LTwo
Holes. D r i l l e d
and t a p p e d ' f o r # 6 .
-=
.126
-=IF
Material: M i l d Steel
N o t e : Make f o u r .
I'
I1
Supply:
1.760
1.650
- 66x1" A l l e n head
Round head
-- Lt6x.260
#6 Washers
16 - 1 6 Lockwashers
4 - 18x.500 Brass
4 - #8 Washers
4 - #8 Lockwashers
8
8
16
4
M a t e r i a l : Aluminum 6 0 6 1
Note: Make f o u r .
Philllps
Measurements: +/-.010
I
n
1 inch
J
0
Mobile Robot Lab
Title:
File:
1984
T r i k e Fork L i m i t Switch Brackets
[rovcr]/usr/ywp/d.triiforkln.dp
Drawn by:
Checked by:
G.Podnar
Dwy.
Fork
L
‘
Y
T u r n l n g Length 42” ( ‘ c u r b t o curb’
0
Mobile Robot Lab
1 inch
Title:
File:
1984
T r i k e Frame
[rovrr]/usr/gwp/d.tri/Frame.dp
Drawn by:
Checked by:
G.Podnar
Dwg*
No.:
Frame
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
00°
00°
0
0
0
0
0
0
0
0
0
0
0
0
Bottom
Lort
Rlght
Part A
N o t e : Make one.
For D e t a l l s see
FrameAl
frameA2
FrameA3
FrarneA4
Trike
Trike
Trike
Trike
0
---
Drawlngs:
Rlght
Top
Left
Bottom
0
Mobile Robot Lab
Material:
1984
T i t l e : T r i k e FrameA
4" S q u a r e Alumlnurn
Tubing.
.125" w a l l .
6063-T6
F i l e : ~rover]/usr/ywp/d.tri/FrameA.dp
Date: 26-Jan-84
II S c a l e : N T S l A.w. r o v e J :
1 fnch
Drawn b y :
Checked by:
G.Podnar
Dwy*
No. :
Frame
n f lil
1.25
N o t e : Same h o l e p a t t e r n
on a l l f o u r s i d e s
and b o t h ends.
i
0
u)
W
I
A1
l.C
7
I
j
0
0
r r i i
I
W
.
N
d
Material:
4" Square A1 uminum
Tublng.
.126" w a l l .
6063-T6
0
z
0
Right
0
Note: D e t a i l o f Frame P a r t A .
N o t e : A l l measurements: + / - . 0 1 6 "
E x c e p t o v e r a l l l e n g t h : +/-.loo"
O r where given.
0
l inch
G8
Tit!e:
0
0
@
Front
\
-Note:
See o t h e r end f o r
these h o l e placements.
I
Mobile Robot Lab
T r l k e FrameAl
-
0
1984
Right
F i l e : [rouer]lusr/gwp/d.tri/FrameAl.dp
D a t e : 26-Jan-84
I S c a l e : NTSl Approved:
Drawn b y : G.Podnar
Checked b y :
NO.:
nf
10
]
N o t e : Same
on a l hl ofloeu pr ast it deer sn
and b o t h ends.
UI
N
Q
r-1.376
\
- 0
9
N
m
2.05
4-
1 l a r g e h o l e ' o n Top
and B o t t o m . 2.76 d l a .
With f o u r small holes
spaced as shown.
D r l l l e d and tapped
f o r #lo s c r e w s .
2.00
M a t e r 1a1 :
4 " Square Aluminum
.125" w a l l .
Tublng.
6063-T6
N
Lo
x
-
2.00
-=I2.00
N o t e : D e t a l l o f Frame P a r t A
N o t e : A l l measurements: + / - . 0 1 6 "
N o t e : Same h o l e p a t t e r n o n
b o t h Top and Bottom.
E x c e p t o v e r a l l l e n g t h : +/-.loo"
O r where g i v e n .
i
0
u!
*
F i l e : [rover]/usr/ywp/d.tri/FraaeA2.dp
S c a l e : N T S l Approved:
Date: 26-Jan-84
Drawn by: G.Podnar
r ame
I
Front
-
N o t e : See o t h e r end f o r
t h e s e h o l e Dlacements.
No.:
F
n f 10
1.50
1.25-
1
1.25
I
N o t e : Same h o l e p a t t e r n
on a l l f o u r s i d e s
and b o t h e n d s .
F
I
I
I
0
0
h
ID
.4
N
9
N
w
t
t
t
Material:
4" Square Aluminum
Tubing.
.125" w a l l .
6063-15
0
Left
0
Note: D e t a l l o f Frame P a r t A.
Note: A l l measurements: +/-.016"
E x c e p t o v e r a l l l e n g t h : +/-.loo"
O r where g l v e n .
0
L
Tit,!:
0
0
@
1984
F i l e : [rover]/usr/gwp/d.tri/FraiiieALI.dp
\
-Note:
Front
0
Mobile Robot Lab
T r i k e FrameA3 - L e f t
See o t h e r end f o r
these h o l e placements.
Date: 26-Jan-84
I S c a l e : N T S l Approved:
O r a w n b y : G.Podnar
Frame
No. :
nf
in
1.25
1.50
N o t e : Same h o l e p a t t e r n
on a l l f o u r s i d e s
and b o t h ends.
-2-cutouts
SS400-PZ M o t o r f o r r e f e r e n c e
:a
-
. -2.00
- -+
- - - 2-. 0 0
-
-r=*
4-hOl e5
\
Two l l n e s marked t o
I n d i c a t e allgnment
f o r welding o f P a r t 8.
r.1.376
0
*
1
-0
1 l a r g e h o l e ' o n Bottom
Top. 2.76 d i a .
Wlth f o u r small hole5
spaced as shown.
D r i l l e d and t a p p e d
f o r Y 1 0 screws.
-
*
Material:
4" Square Aluminum
.125" w a l l .
Tubing.
6063-16
Bottom
N
ID
rd
-
2*oo
-I-2*oo
N o t e : D e t a i l of Frame P a r t A.
-s\
?lf
r(
\Mot@:
Same h o l e p a t t e r n on
b o t h TOD and Bottom.
N o t e : A l l measurements: + / - . 0 1 6 "
Except o v e r a l l l e n g t h : + / - . l o o "
O r where g l v e n .
Mobile Robot Lab
Tlt!e:
T r i k e FrameA4
-
0
1984
Bottom
F i l e : [rover]/usr/gwp/d.tri/FrameA4.dp
Revised:
5-Feb-84
gp
I
D a t e : 26-Jan-84
S c a l e : NTS Approved:
Drawn b y : G.Podnar
,
-Noto:
Front
See o t h e r end f o r
these h o l e placements.
Checked by:
i;:;
Frame
5
of
in
r
2.50
1.26
'
1.26
Part B
4.00
'\\\
L
\
-
i
Front
Right
\ I
I\
\
Back
Left
One l a r g e h o l e on F r o n t
2.76 d l a .
With 4 s m a l l h o l e s
spaced as shown, d r i l l e d
and t a p p e d f o r Y10 screws.
4" Square Alumlnum
Tublng ,
.125" w a l l
.
Chamfer edges a t b o t h ends
f o r w e l d l n g t o P a r t A and
P a r t C.
B
t
\ I/
N o t e : Make ona.
0
Mobile Robot Lab
1984
T i t l e : T r i k e FrameB
.
.
F i1e : [ r o v e r ]/us r / gwp / d t I' i/ F I'anieR d p
J
Rinht
1 inch
Front
I
D a t e : 26-Jan-84
II S c a l e : NTS-0-ved:. .
Drawn by: G.Podnar
Dw9.
No.:
Checked by:
F r ame
nf
ln
2.50
375
N
ID
d
Part C
/
Qottom
Left
lght
One l a r g e h o l e on Top
2.75 d i a . W i t h 4 s m a l l h o l e s
spaced as shown, d r l l l e d
and t a p p e d f o r #lo screws.
Chamfer edges a t b o t h ends
f o r w e l d l n g w l t h P a r t 6 and
P a r t D.
-
Materlal:
4" Square Aluminum
Tublng.
.125" w a l l .
6063-75
N o t e : Make one.
,iBl
c
Mobile Robot Lab
198
T l t l e : T r i k e FrameC
F l l e : rroverl/usr/iiw
- D. / d . t r i / F r a m c C . d p
.
1 Inch
r
Front
I
I
Date: 25-Jan-64
Drawn b y :
Checked by:
I Scale:
G.Podnar
NTS
Approved:
Dwg.
No.:
F raine
nf 10
-k
d-
3.00
3.00
4-
3.00
4
-
<I;->
IO
I
7-1- -
Two l i n e s m a r k e d t o
l n d l c a t e alignment
for weldlng o f P a r t C.
-
,
$--
/'
I
I
Part D
6.
I
I\
Front
TOP
1 0 - h o l e s l o p and E o t t o m . 3 7 5
@I
0
Mobile Robot Lab
1984
T i t l e : T r i k e FrameD
F i l e : [rover]/u~r/gwp/d.trl/FrameD.dp
1 inch
Front
L
.
1
I
_
~
D a t e . 20-Jan-84
S c a l e : NTS
Drawn by: G.Podnar
Approved:
Checked b y :
No. :
.
_
_
I
_
-
Frame
of
in
* .125
P
Note
Eight holes
d r i l l e d t o a depth
of . 7 5 " f r o m s u r f a c e
and t a p p e d f o r #la screws
Note
One h o l e , c e n t e r e d
1 . 0 0 " d 1 a.
B e v e l e d f r o n t and b a c k .
Note: Four h o l e s . 2 7 5 " dla.
A l l spaced e v e n l y
f r o m c e n t e r on a
2.5" square.
M a t e r i a l : Alumlnun
Note: A l l measurements: +/-.010"
O r where g l v e n .
N o t e : Make Two.
0
Mobile Robot Lab
T i t l e : T r i k e Frame " A "
1984
End C a p s
F i l e : -p>ver]/usr/gwp/d.tri/FranleCapA.dp
D a t e : 27-Jan-84
S c a l e : NTS Approved:
I
I-
DldWn b y :
Inch Checked b y :
'
I
G.Podnar
Dwg.
No. :
Frame
nf
In
I
I
I
I
I
I
I
I
I
I
I
I
I
I
1
I
I
I
I
I
I
I
I
I
I
1
I
I
I
I
I
I
I
I
I
I
I
I
I
I
.I
I
I
I
I
I
I
I
I
I
I
1
I
I
I
I
I
I
I
I
I
I
I
I
1
I
I
I
I
I
I
I
I
1
Two h o l e s . 3 7 5 " d i a .
D r i l l e d Through
li
A l l bevels
-x
.loo
46'
RPB 6/8
for reference.
M a t e r i a l : Aluminum
N o t e : A l l measurements: + / - . 0 1 0 "
Or w h e r e g i v e n .
N o t e : Wake Two.
1,z4
Mobile Robot Lab
T i t l e : T r i k e Frame "D" End Caps
File: [rover]/usr/gwp/d.tri/Fr~me~~~)D.dp
D a t e : 27-Jan-84
IS c a l e : NTSl Approved:
Drawn by: G.Podnar
rame
No. :
r
nf
~n
m
~
Right
Materlal:
4" Square Aluminum
Tublng. . 1 2 6 " wall.
6063-T6
Front
I
T
c
0
0
1.25
---
+
3.00
1.26
0
'*'o
*
Material:
1 Inch
I
L
Four h o l e s
.210" d i a .
.0625" Aluminum
N o t e : Wake t e n .
GI
0
Mobile Robot Lab
1981
T i t l e : T r i k e Frame C o v e r s
.
F i1 e : [ r o v e r 3 /us r / gwp / (1, t r i/ F r aine C o v dp
S c a l e : NlS( Approved:
D a t e : 17-Feb-84
I
Drawn by:
Checked by:
I
G.Podnar
Dwg*
No.:
F r ame
f
Two H o l e s .
D r i l l e d and t a p p o d
f o r #lo-32
0
-7
r
= a (May b e f i l e d o r g r o u n d )
.226
N o t e : A l l measurements: + / - . 0 1 6 "
E x c e p t H o l e s p a c i n g : +/-.010"
Or w h e r e g l v e n .
N o t e : Make t w o .
Material: Mlld Steel
Supply: 4 each:
fl1
10-32 A l l e n Cap S c r e w s
Washers
I . o c k Washers
Mobile Robot Lab
T i t l e : T r i k e Frame S t e e r i n a Motor SlJooort.
File: [rover]/usr/gwp/d.tr~/F~a~~SupA.dp
Drawn b y :
N o t e : S e e o t h e r e n d for
these h o l e placements.
I
_
G.Podnar
Dwg. Frame
,:I
Hitch Tongue
p.750
Hitch Bracket
4
J
I
I
One h o l e .
.375 dia.
7 7
1.250
0
0
I
-
.
E
\
I
IIr-
.200 max
inslde radius
IL
.la75
1.250
-
1
I
I
4
I
One H o l e .
.E60 d l a .
0
I
9
0
I
0
I
4
Materlal: Mlld Steel
Note: Make one o f each.
9
1-
1
I
Mobile Robot Lab
\
Four H o l e s
.375 d f a .
Checked b v :
_I__-
-
D W Q . Frame
No.:
of
+io
4.00
\
I4
Four h o l e s . 3 7 5 d i a .
.60
Front
Material: Mild Steel
N o t e : Make f i v e .
Measurements: + / - . 0 2 0
0
Mobile Robot Lab
198.1
T i t l e : T r i k e Frame W e i g h t s
( A b o u t 9 pounds e a c h . )
C h a r g e : 1-11921
t F i l e : [rover]/usr/ywp/d.tri/Fra~~~eW.Op
D a t e : 24-Mar-84
S c a l e : NlS Approved:
I
1 inch
U
brawn by:
Checked by:
G.Podnar
Dwg.
No.:
Frame
f
Appendix I1
Appendix I1
Electrical Drawings
Appendix I1
110 AC
S t e e r l n g Motor
Superlor E l e c t r l c
S1 o-Syn S S 4 0 0
I-iiTFij
Opto-22 DC2OOP
e
IC
k o p i a n Model #U9YlOOO
iVOC U n r e g u l a t e d ( 1 0 A 86)
110 AC
ID-AC-J
Opto-22
240D10
0
I
t
I
I
.
\
I
3
I
/
I
D r l v e Motor
Superlor E l e c t r l c
S l o - s y n SS1800
H o l d i n g Torque Requirements
VD C
Current
Toraue
R e s i d u a l H-T
S l o - S y n SS400
30 VDC
0.60 A
600 or-In
9 oz-In
S l o - s y n SS1800
9.6 VDC
4.7
2200 o z - I n
20 o r - I n
Mobile Robot Lab
Title:
A
N o t e : The g e a r l n q o n t h e S t e e r l n g m o t o r w l l l be h i g h enough t h a t
DC h o l d i n g t o r q u e w o n ' t b e n e c e s s a r y and l i t t l e o r no
e x t e r n a l f o r c e s on t h e wheel I t s e l f a r e e x p e c t e d .
Neptune r o v e r e l e c t r s n r c s
I C h e c k e d by:
I.".
.
0
198.1
11OVAC
Power C o n n e c t o r
I
I
I
Hot Q
DC H o l d i n g et,(
Calex 1.5.500
Neutral
AC
Q
Cmn
.
I
%
I-Ea-L
I I
I
DC 2OOP
C h a s s i s Ground
AL
stop
5~ ? o r r e s t o f c o n t r o l l e r
stop
A d d i t l o n a l P a n i c B u t t o n s a r e I n s e r l e s w l t h t h e one shown
litle:
Emergency S t a r t / S t . o p s y s t e m
F i l e : rroverl/usrlnivek/rJver/ne~tune/mreZ.dp
Date:
3/6/84
I S c a l e : NTS Approved:
nws.
Drawn b y :
nivek
"2.:
Checked b y :
SIDE]
'DE THREE
-
Power Supply
1.5.500
K
I
4.00 inches
16.100
>I, &
o
r
DC SSR
and
-
I
SIDE TWO
SIDE TWO
v2
SIDE FOUR
SIDE THREE
SIDE THREE
~~
G51
T I t 1e :
SIDE FOUR
~
SIDE ONE
SIDE ONE
Mobile Robot Lab
0
190
-
N e p t u n e ' r o v e r e l e c t r o n Ics
F l e c t r o n i c s Box 1 a y o u t Assembled V i e r
[rover]/usr/nlvek/rover/ncptune/n1re3.dp
2/9/84
( S c a l e : NTS Approved:
Flle:
Date:
Drawn b y :
Checked b y :
nlvek
-
Dwg
.
No. :
main c i r c u i t b r e a k e r
CAM1
START
processor connect
I
CAM2
V deo o u t
7
K.-
1
3.0 i n c h e s
E.6
..
I
see c i r c u i t b r e a k e r
1
i n c h 3
1.626 i n c h e s diam.
1.26 I n c h e s
1.5 i n c h e s
K
2.0 i n c h e s
1.1875 i n c h e s diam.
K
2.6 i n c h e s
.XI12 Inches
>It-
1
-
4.75 i n c h e s
C u t o u t Diagram
SIDE TWO
S I D E TWO
HREEioi
H
one i n c h
SIDE THRFE
S I D E FOI
SIDE ONE
~~
SIDE ONE
View f r o m t h i s s i d e
E l e c t r i c a l Box Sides
!/9/84
[ C h eDrawn
c k e d by:
by:
.
SIIIE ONE
I S c a l e : N T S l Approved:
Dwg
nivek
..
.
.
rw
. :
0
1
2
3
0
0
0
0
1
0
0
1
1
0
0
1
0
1
0
1
0
1
1
1
0
1
1
1
NOT( 1 A I D 2 )
Legal s t a t e s
Legal s t a t e s are
I l l e g a l states
S w l t c h e s 1 and 2 c a n n o t b e o n a t t h e same t i m e
1fi1
litle:
Mobile Robot Lab
L i m i t switch band placement
Drawn by:
n4Vt.k
No.:
0
1994
@*
0.20 I n c h e s w/ c l e a r a n c e
2.02 I n c h e s d i a . c u t o u t
/
2.50 Inch dla. b o l t c i r c l e
2"
0
\
The h o l e s a r e f o r m o u n t l n g t h e s e c o n n e c t o r s
SIDE TWO
Mobile Robot Lab
E FOUR
S I D E FOUR
S I D E THREE
SIDE THR
S I D E ONE
SIDE ONE
Title:
Neptune r o v e r e l e c t r o n l c s
SIDE TWO
E l e c t r o n i c s Box l a y o u t
Date:
2/9/64
Orawn by:
Checked by:
I S c a l e : NTS Approved:
Dwg
nivek
.
No. :
0
1984
K.-
3.625 Inches
1-
K-
1-1
3.625 inches
-
1.50 i n c h e s
h
7
k
5 inch*
8
8
7
&
5 i n c h d
8
"T
I
inch]
8
@
1.876 inches
L
0
@
T
1.376 Inches
d
IDE THREE
k
7
6 I n c h d
3.00 inches
The box b o t t o m I s 1 2 i n c h e s s q u a r e
View f r o m t h e t o p
A l l h o l e s a r e f o r 8-32 screws w i t h c l e a r a n c e .
The i n n e r s q u a r e i s t h e 1 1 2 i n c h l i p a r o u n d t h e t o p .
FaE
S I D l TWO
Iml
FOUR
SIDE FOUR
SIDE THREE
SIDE THREE
Mobile Robot Lab
~
SIDE ONE
SIDE ONE
--
Date:
2/9/94
Drawn by:
Checked b y :
! S c a l e : fJTS Approved:
Dwg .
nivek
No.:
0
I1984
0
0
0
0
( S I D E j
one I n c h
one i n c h
0
H o l e s a r e f o r 10-32 Screws w i t h c l e a r a n c s
lfil
S I D E TWO
Mobile Robot Lab
FOUR
SIDE FOUR
SIDE THREE
S I D E THREE
SIDE ONE
SIDE ONE
D a t e : 3/5/84
Orawn by:
Checked b y :
I S c a l e : NTS Approved:
Dwg
nivek
No. :
.
0
1984
>
0
C e n t e r h o l e i n p l a t e t o be threaded f o r t h e s w l t c h .
N o t sure about t h r e a d l n g d l a m e t e r r .
Mobile Robot Lab
No. :
Q
1984
4.7 K
1 K
T h i s symbol w i
Date:
3/6/84
Drawn b y :
Checked b y :
I S c a l e : NTS Approved:
Dwg
nlvek
NO. :
.
computer
computer
Manual C o n t r o l l e r
ls e l ec
S-AC-1
+
S-AC-2
+
Delay
Steer
+
D-AC-2
+
Delay
+
D-AC-1
+
Delay
Drive
-
D-DC-1
+
Delay
I f d r i v e I s on t h e n t u r n o f f DC
holding torque
Mobile Robot Lab
Title:
When e l t h e r o f t h e manual c o n t r o l l e r s w i t c h e s a r e moved i n
e i t h e r d i r e c t l o n t h e S e l e c t l l n e on t h e Mux i n p u t s i s p u l l e d
l o w , t h u s d i s a b l i n g computer c o n t r o l .
0
1984
Neptune r o v e r e l e c t r o n l c s
C i r c u i t diagrams
F i l e : [roverl/usr/nivek/rover/neptune/mrel1.dp
Date:
3/7/84
I S c a l e : NTS Approved:
Drawn by:
nivek
Dwg
No. :
Checked by:
.
Coax R e l a y
16VDC
760 ohms
status lights
-+I
Camera 1
-
I
I
I
Ln n
Camera 2
computer
Manual
5 VDC
-
'Title:
neptune rover e l e c t r o n i c s
camera s w i t c h i n g c i r c u i t r y
F i l e : ~rover~/usr/nivek/rover/neptune/mrel2.dp
Date:
3/8/84
I S c a l e : NTS Approved:
nwg
Drawn by:
nivek
No. :
Checked b y :
.
,
Appendix 111
Appendix 1x1
On-Board Processor Software
A simple controller for the Neptune rovcr.
This program parses commands from the VAX or the debugging terminal and translates them in to robot
motion and control. On start up, the debugging terminal is queried for the source of commands: either the
VAX, or the debugging terminal for manual control. When an entire command has been parsed, the
commands arc executed. At the end of execution, a ’!’ is sent to the terminal, and also to the VAX, if it is
originating commands. If there was a problem with the parse, a ?’ is sent and nothing is done.
Commands are in the form:
[ subcommand subcommand subcommand
... ]
Subcommands are either a single character or a movement command enclosed in parenthesis.
The single character subcominands are:
C : C e n t e r s t e e r i n g motor
0 : S e l e c t camera z e r o ( l e f t )
1 : S e l e c t camera one ( r i g h t )
The movement subcommand is of the form:
(motion. d u r a t i o n )
where motion is one or two characters:
N :
:
L :
R :
F :
B :
C
Do nothing
Center s t e e r i n g
Steer left
Steer right
Drive forward
Drive backwards
and duration is a decimal number representing the number of 1/6Oths of a second to run this particular
motion. The N command is mutually exclusive of everything else, the C, L and R commands are mutually
exclusive, as are the F and B commands. If you specify two or more mutually exclusive commands, the parser
will return an error.
In all commands, upper and lower case is identical and spaces, carriage returns and line feeds are ignored.
If the out-of-bounds limit switch is triggered during movement, all steering in the out-of-bounds direction is
inhibited, but execution continues.
Here‘s an example command:
[ 0 C (LF, 60) (CF,
120) ( R B , 60) ( 6 , 120)
3
This command would be completely parsed, and then the following actions would occur: The left camera will
be selected, the steering motor will be centcred, the robot will move forward steering left for one second,
forward stceriiig to the ccntcr position for two sxonds, backwards steering right for one second, and
backwards for two seconds. After all of this had been performed, a ’!’ would be sent to the terminal, and the
VAX, if it were issuing commands.
While running, a rK typed at the terminal will immediately abort the current command and turn off all of the
motors.