SPE 111125

The Evolution of Automation in Drilling

Alfred W. Eustes III, Colorado School of Mines

Copyright 2007, Society of Petroleum Engineers

This paper was prepared for presentation at the 2007 SPE Annual Technical Conference and
Exhibition held in Anaheim, California, U.S.A., 1114 November 2007.
This paper was selected for presentation by an SPE Program Committee following review of
information contained in an abstract submitted by the author(s). Contents of the paper, as
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at
SPE meetings are subject to publication review by Editorial Committees of the Society of
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
for commercial purposes without the written consent of the Society of Petroleum Engineers is
prohibited. Permission to reproduce in print is restricted to an abstract of not more than
300 words; illustrations may not be copied. The abstract must contain conspicuous
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, Texas 75083-3836 U.S.A., fax 01-972-952-9435.

Abstract
This paper has been written for the keynote speech to be given
in the Drilling Automation Where Are The Game
Changers? plenary session for the 2007 SPE Annual
Technical Conference and Exhibition in Anaheim, California.
The paper covers the evolution of drilling mechanization and
automation from the mid-nineteenth century to today. In the
history of drilling, there have many aspects of drilling
mechanization and automation that have been investigated,
machines designed and built, and operations implemented
with varying levels of success and many failures. These
aspects include drilling fluid systems, cementing operations,
downhole automation, and the rig floor. Because drilling
mechanization and automation history has been centered on
the rig floor and related surface operations, the focus of this
paper is on the rig floor for drilling operations such as rate of
penetration, rotary systems, rig floor mechanization and
automation, and entire drilling rig automation.
Introduction
Mankind has always found wealth under the Earths surface,
whether it be water, brine, oil, gas, or gems and minerals.
Accessing that wealth has been the challenge. The drilling rig
is one of the machines developed to access that wealth.
It all started with a pole, a rope, a heavy weight, and a lot
of labor. These first rigs were percussive drilling units called
cable tool rigs. These units, popular until the mid-twentieth
century, were the only way to drill until the mid-nineteenth
century. On January 9, 1845, Robert Beart in Great Britain
was granted a patent on the first of a new style of rig, the
rotary rig with continuous circulation.5, 8
Regardless of whether percussive or rotary systems were
used in these early years, the tasks were repetitive and
required human manual strength. These early rigs were a
nightmare of dangerous and unguarded equipment and
hazardous operations. The roughneck was not named thus for
his knowledge.

Anywhere repetitive or dangerous operations take place,

the allure of mechanization and autonomy beckons. The
automotive and aeronautical industries have recognized this
for years. The automotive industry in particular has
mechanized and automated the manufacturing floor and leads
in many areas of robotics. The drilling rig floor is clearly one
of those places where this can be useful.3
The process for technology development follows a path.
Phil Vollands of Varco International has stated that there are
three eras of evolution: mechanization, semi-automation, and
local automation.12 The process starts with mechanization.
That is, removing human power and labor and substituting
mechanical power. Not only does that get the human out of
the way, it allows for more force and torque to be applied.
The next step is to automate the particular operation. At this
point, the humans role changes from brute strength to using
their intelligence to supply the machine the brains and vision
for the automated operation. That makes this a local semiautonomous operation. An ultimate goal is to completely
automate an operation making it totally autonomous. No
human intervention is needed except for startup and for
emergencies. This is becoming a reality with the advent of
robust computers and programmable logic circuits (PLC).
The motivation for rig mechanization and developing
autonomous rigs are many. It starts with safety. Removing
people from the area of heavy moving machinery clearly
enhances safety. This is the goal of many operators,
contractors, service companies, and regulatory agencies
around the world.
Another motive is to reduce the number of people on the
rig floor. Over the last few decades, the number of people
entering the drilling industry has been declining.
Mechanization and automation can reduce the required
number of people not only on the rig floor, but also on the rig
itself. It also reduces costs associated with the head count.
Operating in harsh environments is another motive. In
cold and windy or severe wave conditions, human ability to
control the rig floor diminishes. Machines designed for the
particular environment are not as affected and can continue
operations whereas human operations would have to wait on
weather.
There are also efficiency motives.
For example,
efficiencies manifest themselves in the area of drilling
operations where the optimum rate-of-penetration can be
determined and operations set at that condition. Operational
efficiencies can be gained with the automation of repetitive
tasks such as tripping. The tripping operation can be faster,
safer, and more precise. Tripping speeds can be more precise

and closer to computed maximums to minimize surge and

swab problems. In addition, since a machine can be
programmed, more complex operations can take place.
Another motive for automation includes reducing rig weight
and size allowing for a smaller rig with a smaller
environmental footprint. Smaller rigs also mean reducing
mobilization and demobilization time. A rig that is automated
has the potential to be remotely monitored and controlled,
literally from around the world.
And finally, from the
beginning of the history of rig mechanization and automation,
one of the primary motives has been to reduce operating
costs.2, 3, 10, 12, 15, 20, 21
Automating Drilling Operations
Automated drilling operations started with the rock
penetration process itself. That is, the feed control of the bit
as it drills. Leonardo DiVinci developed a drill feed using a
screw mechanism back in the 1500s. However, there is no
evidence it was actually built.8
The first construction of an automatic feed control is from
the early 1860s. Rodolphe Leschot of France built the first
automatic bit feed for his invention of the diamond core rig for
drilling blast holes for European tunnels. Leschots drilling
system consisted of a hydraulic piston that could push and
rotate the bit. The hydraulic piston was fed by a steam driven
water pump. The unit could push down when there was not
enough drill string weight. When the drill string weight was
sufficient, a valve was opened just enough to regulate the feed
rate into the hole.3, 8
It wasnt until the early twentieth century that drillers even
knew what the weight on bit was. The earliest patent on a bit
weight indicator was granted in 1906 to John Sharp, the
brother of Walter of drill bit fame. Howard Hughes Sr. also
received patents on various bit weight indicators between
1914 and 1923. However, these were not particularly
sucessful.9
The first weight indicator to look like todays systems was
developed by Elmer Decker and Frosty Martin in 1927. It had
a piston fixed at the vertex of a triangle with the drilling line
pushing on the piston. The deflection of the drilling line was
a direct measurement of the load. However, this invention
didnt work very well as it leaked. It needed the development
of O-rings, which had not been invented yet. A young
engineer offhandedly suggested using a diaphragm instead.
Thus an industry was born.9
It still took the development of the drawworks brake
before the automatic drill feed could be practical. Prior to that
time, most drilling feeds were controlled by a single band
brake on a cast iron brake drum. Most drill feed systems in
the early twentieth century made use of a feed mechanism that
employed the band brake, not unlike some of today. The
metal to metal brakes of 1912 did not allow for a smooth feed.
Improved brake linings and brake rims helped this situation
somewhat. The hydrodynamic brake was developed in the
1930s and the electromagnetic brake later that decade.8
There were various approaches to automatic drilling feed
mechanisms. Paraphrasing Brantly from his book, there was
the old tried and true driller with his hand on the brake; the use
hydraulic cylinder feeds, similar to Leschotes mechanism; the

111125

use of drilling torque or rotational power; and a weight on bit

control system.8
In the late 1920s, Halliburton (National Supply Company)
and Hild (Oil Well Supply Company) both built a torque
based machine. The Halliburton mechanism consisted of a
steam or electric motor with a differential gear on a jack shaft.
The engine was on another jack shaft. When the torque to
drill exceeded a certain amount, the drill string was retrieved
until the torque dropped back to the prescribed condition. The
Hild mechanism was similar but used an electric rotary and
drill feed motor.8
In the 1930s, hydraulic feed rotary tables were developed.
The Sheldon machine consisted of two hydraulic cylinders on
either side of the rotary table. It could be used conventionally
with a kelly or as a 30 inch (76 cm) stroke hydraulic system.
The Doheny-Stone Company (predecessor to Hydril) came out
with a hydraulic feed system, too. It consisted of threecylinders and a hoist. These units, along with others of this
time, had stroke lengths of 2 10 feet (0.6 3 m). This
restriction, along with the fact that they were slow and came
out in the economically disastrous Great Depression, limited
their deployment.3, 8
The weight-on-bit automatic drill feeds were what
eventually took the market. According to an Oil and Gas
Journal article from 1928, a mechanical differential was used
to regulate and stabilize the pressure of the drill stem on the
bit. In other words, this is weight-on-bit. These systems
were chain driven and set between the prime movers and the
drawworks. They were connected to two electric motors, a 7.5
hp (5 kW) motor for drilling and a 35 hp (26 kW) regulating
motor. It was this motor that took up the slack from the
drilling motor when the bit loaded excessively.16
In 1935, Dillon, Dreyer, and Jenks of Westinghouse
patented an automatic driller for rotary systems. This system
was based on managing the weight (or pressure) on the bit
through an electro-mechanical device.
In the 1940s,
pneumatic actuated feed control of band brakes were designed
and built.3, 11
Regardless of the type, these mechanical systems had
limitations of preciseness and non-linear response curves.
This made controlling the rigs brakes challenging. With the
advent of new types of drawworks brakes and of computer
control, new automatic drill feed systems were developed. In
addition, further refinements in rig instrumentation and
analysis of drilling variables have helped push these systems
further.7
Even as early as 1971, the use of computer controls for rig
operations was being considered. One such unit was by
Baroid. The digital/analog system that included a mixture of
electrical and pneumatic components, monitored variables
such as depth and time along with normal mud log type data.
The control system was a closed loop that measured and
changed weight-on-bit and rotary speed (and secondarily
torque) by varying the engine throttle. The system used
simulations to compare actual versus expected conditions and
made changes accordingly.15
In 1997, Helmerich and Payne and Varco develop an
electronic bit-feed control system. This effort resulted in
significant improvements in drilling efficiency. However, as
is true with any computer system, it is only as good as the

111125

software. In addition, it took a well trained driller to operate

the system. And if anything went wrong, it was impossible to
solve in the field.3, 6
Now, the systems in use not only can control weight-onbit, but also can maintain rate-of-penetration, constant
hydraulic pressure, and torque. These systems are also
capable of handling specific operations ranging from
automated tripping and reaming, drag and torque
measurements, pump start, continuous bit drill-off testing, and
stick-slip prevention. These systems have reduced the need
of a drillers proficiency in computer programming. And the
drillers consoles have gone from rudimentary weight and
pressure gauges standing out in the weather to sophisticated
computer screens with process specific information (i.e. auto
drilling, tripping, etc.), joystick controls, and comfortable
chairs in a climate controlled doghouse.2, 7, 13, 20, 21
Automating the Rotary System
Another invention that helped in the automation of drilling
systems is the top drive. However, back before mid-1980, it
was known as a power swivel (or a power sub). Previous to
this time, power swivels were typically used in workovers
only. As will be noted, drilling operations with power swivels
were a relatively rare event.
In 1955, Paul Scott in Michigan developed the first
hydraulic power swivel and hoist. The PS-200 model built in
1958 was the follow on design that is the basis of more
modern designs. This unit was originally designed for coring
operations, taking over for cable tools in the Michigan basin.
These early model power swivels eventually went on to
drilling operations with up to 15 inch (0.46 m) diameter holes
in glacially deposited material. These operations drilled
75,000 feet (22,860 m) of 6 1/4 inch (15.6 cm) to 7 7/8 inch
(20 cm) holes in North Louisiana.3, 4
Offshore, power swivels were in use in the late 1950s. A
power swivel was used on the NOLA 1 drillship. In the early
1960s, other hydraulically operated power swivels were in
use on the Shells Eureka core ship and the Glomar
Challenger.23
In the early 1970s, Brown Oil Tool and Bowen developed
the first electric power swivel for ARCO Oil and Gas. In
1983, a Varco electric power swivel was designed and placed
on two SEDCO jackup rigs. This swivel was called a Top
Drive Drilling System. It had a 1,000 hp (750 kW) DC
motor and pipe-handling system. What made this unit so
successful was the integration of pipe handling equipment to
make and break connections built into the power swivel
system. 3, 23
Now, the top drive unit is ubiquitous offshore and is
infiltrating onshore rigs in greater numbers.
Automating the Rig Floor
The rig floor has seen the most automation of all drilling
systems. Most of the danger on a rig exists on the rig floor.
Moving people off the floor enhances safety significantly. All
new offshore drilling vessels have incorporated rig
mechanization and remotely controlled pipe handling
systems.2
The first attempts at mechanizing the rig floor involved
using chain tongs, spinning chains, and breakout tongs. The

first powered rig floor system consisted of making slips easier

to handle. In 1945, the Byron Jackson Company introduced
the BJ Power Slip, a pneumatically controlled set of slips. The
slip operator had a foot pedal that controlled the opening and
closing of the slips. These slips could be easily changed and
were full opening. Varco introduced a similar set of power
slips in 1950. In 1955, National Supply Company built a
rotary table that integrated power slips into the design.1, 8
Perhaps the most important mechanization improvement to
the rig floor was the iron roughneck. In 1975, the first
mechanical iron roughnecks were developed by Varco. A
1977 patent claims that the iron roughneck is a combination of
power slip, spinner, and torque wrench in a moveable mount.
Since that time, iron roughnecks have been diversified to
include handling not only drill pipe, but also drill collars,
casing, bottom hole assemblies including stabilizers, and
tubing. The latest iron roughnecks have included remotely
operated mud buckets, cleaning and doping system, and
stabbing guides.
They can also handle the torque
requirements of the latest generation of pipe threads.2, 6
However, once the drill string is unscrewed, it has to be
either set back or laid down. Racking systems are another part
of the rig floor automation development. In 1949, Byron
Jackson Company developed a prototype of a three-arm pipe
racking system. By 1956, the first floating rig included a laydown pipe-racker and horizontal pipe storage system. In
1974, a Sedco drillship had the first Western Gear horizontal
pipe racking system. By 1981, the first semi-submersible rigs
included a mechanical racking system. And in 1986, the first
full column racking system was introduced on the Transocean
8 semi-submersible rig.3
What have made pipe racking systems viable are computer
controllers. Until the introduction of computer systems,
humans were needed to control the hydraulic systems through
joystick control. This was difficult at best. With the
introduction of flash memory (impervious to vibration
damage) and of simple programming techniques, the
movement of pipe could be preset to various locations without
the constant pipe positioning with the joystick.20
Varco introduced a pipe pick-up and lay-down system that
completely eliminated human intervention anywhere in the
pipe path. In 1993, the first system to remotely manage pipe
on the pipe deck was developed. By 1996, the first modular
pipe-racking system was installed. The process continues with
each addition to the offshore rig fleet.3, 20
Automating the Entire Rig
The first plans to put an entirely automatic drilling rig together
were in the late 1940s. Plans and drawings were completed;
but, the rig was never built. It consisted of a mechanized
hydraulic elevator with the hydraulic cylinders set below
ground level (in a mouse hole type of arrangement). An
electric rotary was placed on top of these hydraulic cylinders
riding up and down, rather like a top drive. The unit was a
single lay-down type that automatically racked the pipe.8
A rotary rig with hydraulic lifts rather than a mast was
shown at the 1954 Tulsa Oil Show. However, there is no
evidence that it was actually used in the field.8
One advantage of this type of rig is that it has a smaller
floor footprint and a lower center of gravity. This made sense

for offshore operations. In the early 1950s, the first drillships

included a pipe racking system and used power swivel and
power subs.3
In 1966, the SPE held an automation symposium in Hobbs,
NM. One paper presented was the semi-automatic drilling rig.
This rig was designed for the US Atomic Energy Commission
for post-explosion penetration and sampling of the chimneys
developed from underground nuclear explosions. The rig had
a hydraulic power package. The hoist was powered by
hydraulics; and, the rig had a remote-controlled, trackmounted, power swivel; the forerunner of the top drive. It had
hydraulic power tongs, backup tongs, and pipe-handling
equipment along with pneumatic air slips. The rig was
designed to be crewed with two men and to be able to drill
shallow holes. This rig consisted of two trailers, one with the
power pack and one with the mast that could be slanted to 30
degrees from vertical. This rig was also used in Michigan,
South Texas, and North Louisiana for petroleum related
activities.4
Another automatic rig was outlined in a May 1970 World
Oil magazine article. Bandera Drilling built an automated
drill rig designed to handle 4-1/2 inch (11.4 cm) drill pipe to
13,000 feet (3,960 m). The design criteria also included a
small footprint of 35 foot (10.7 m) by 48 foot (14.6 m) with
racks and a total weight of 300,000 lb (136,000 kg). The rig
was designed for completed automated trips that were
controlled from a van 40 feet (12.2 m) away from the floor. It
had a hydraulic drill head that could lift and lower by 35
feet on a 42 foot (12.8 m) tower. A hydraulic chuck was used
to make and break connections with an electronic-hydraulic
slip. There was a transfer arm that would rotate vertical pipe
to horizontal and would place the pipe in a storage bin and
vice-versa. The hydraulic system was powered by five 350 hp
(261 kW) diesel engines powering a like number of 325 hp
(242 kW) hydraulic pumps.10
As noted in the above article however, the rig did not have
the capability to observe and react to kicks, needed onsite
maintenance and had to change bits manually. People were
still needed on the floor.
None-the-less, these were attempts at developing the fully
automated drill rig. The process continues with todays rigs
becoming more automated, especially in the offshore drilling
arena.
Onshore, examples of the process include the
Helmerich and Paines Flex series of rigs and Nabors PACE
rigs. The inclusion of robust computers has advanced the
science of rig automation. The automation process has
continued with the various rig equipment beginning to talk
with each other in concert. However, this automation is what
is called, local. That is, the automation is instigated locally,
either in the doghouse or the rig floor.3
With the advent of satellite communications, the time has
come for not only remote monitoring but even the possibility
of remote control. In 2004, Schlumberger and M/D Totco
demonstrated that they could remotely control a drilling
operation. From Schlumbergers Cambridge Research Center
in England, they commanded rig functions (such as WOB,
flow rate, rotary speed) at the Cameron Test Facility in
Texas.3

111125

Other Areas of Drilling Automation

Although this paper is concentrated on the automation of
drilling operations and the rig floor, other areas have been
mechanized and automated. For example, a mud mixing
system by IMCO services was developed and implemented on
the No. 1 Bertha Rogers well, a former U.S. well depth record
holder located in Oklahoma. The mud system monitored the
mud weight going in the hole and automatically added barite
as needed to maintain a given weight.14
Cementing operations have been enhanced by automation.
Recirculating cement mixers are an example of this trend.
With a recirculating system, a controller compares measured
cement density leaving the mixer to the desired density and
adds or removes water and/or cement, adjusts pump rate, and
can recirculate the slurry back to the mixer until the numbers
match. This allows for the precise cement density to be
pumped downhole.17
Another drilling aspect recently to become semiautonomous is the bottom hole assembly. Since the mid-1970,
MWD, LWD, and PWD systems have continually increased
the amount of real-time downhole data available to the
surface. This information has allowed more direct control of
trajectories into optimal areas of the reservoir. In the mid1990s, rotary steerable systems were developed. Since then,
rotary steerable systems have evolved into machines that can
guide themselves, with human direction, in various
trajectories.3
Conclusions
The evolution of mechanization and automation in drilling has
been a long process. It started with Leschots auto-drill feed
mechanism in the early 1860s. This trend has been
accelerating, especially with the introduction of computers to
the rig. From that small beginning in the mid-nineteenth
century, the drilling industry has been pushing to mechanize
and automate drilling operations. The primary motives are to:

Increase rig safety, especially on rig floors

Reduce the number of people on the rig floor and on
the rig
Be able to continue operating in harsh weather
conditions
Improve drilling rate-of-penetration
Become more time efficient for drilling operations
such as tripping
Add precision to potentially complex drilling
operations
Make the rig smaller to weigh less and reduce
environmental footprint
Be able to mobilize and demobilize the rig faster
Potentially have remote operations
And the number one motive, to reduce drilling costs

These are some of the various motives that drive many

operators, contractors, service companies, and regulatory
agencies.
There are some in the business that feel humans will
always have the upper hand in all rig operations. Others feel
that full rig automation may help with upcoming personnel

111125

shortages.
A wide variety of opinions exist. In the
January/February 2007 article in the Drilling Contractor
magazine, thirteen drilling and completion experts give their
opinions regarding the future of rigs and wells. These
opinions run from remotely controlled operations to less
automation on the rig floor.16
Whichever way this drilling industry goes, it will prove to
be interesting.
Acknowledgements
I want to acknowledge a number of individuals that helped in
the preparation of this document. The input of Fred Florence,
National Oilwell Varco; Walt Aldred, Schlumberger; and
William Maurer, Maurer Enterprises is greatly appreciated. In
addition, Mary Dimataris, M-I SWACO, was of great
assistance is collecting many of the references. I also
appreciate Mario Zamora, M-I SWACO, for his faith and
assistance in the preparation of this document. Finally, I
appreciate the support of the Colorado School of Mines and of
my wife.
References
1. BJ Services History:
http://www.bjservices.com/website/indx.nsf/WebPages/History?
OpenDocument, last accessed 24 July 2007.
2. Abrahamsen, E., Mechanized Rig Technology Increases Safety,
Efficiency:
http://www.scandoil.com/moxie_issue/issue_78/2005_7-8/mechanized-rig-technology.shtml, last accessed 26
Jul 2007.
3. Aldred, W. et al., "Changing the Way We Drill". Oilfield Review,
17(1): Spring 2005, 42 - 49.
4. Allen, H.G. and Scott, P., "Semi-Automatic Drilling Rig" SPE
1378, presented at SPE Automation Symposium, Hobbs, NM, 28
- 29 April 1966.
5. ASME, Industrial Resources and Equipment - History Resources:
http://www.asme.org/Communities/History/Resources/Industrial_
Resources.cfm, last accessed 24 July 2007.
6. Boyadjieff, G., 'Tool for connecting and disconnection well pipe",
4,023,449, 17 May 1977.
7. Boyadjieff, G., Murray, D., Orr, A., Porche, M. and Thompson,
P., "Design Considerations and Field Performance of an
Advanced Automatic Driller" SPE/IADC 79827, presented at
SPE/IADC Drilling Conference, Amsterdam, NL, 19 - 21
February 2003.
8. Brantly, J.E., History of Oil Well Drilling. Gulf Publishing,
Houston, 1971.
9. Decker, E.L., "Birth of the Bit Weight Indicator". Petroleum
Engineer International, 51(10): August 1979, 80.
10. Dillard, M.M., "New Automatic Rig Improves Drilling Rates".
World Oil, 170(6): May 1970, 87 - 90.
11. Dillon, H.G., Dreyer, W.C. and Jenks, H.C., 'Automatic Drilling
System for Rotary Drilling Equipment", 2,005,889, 12 November
1932.
12. Gaddy, D., "Remote-controlled operations to benefit drilling
industry". Oil and Gas Journal, 99(1): 1 January 2001.
13. Harbour, W.D., Kracik, J. and Ford, D., "An Ergonomic, Process
Oriented Approach to Driller's Consoles" IADC/SPE 39330,
presented at 1998 IADC/SPE Drilling Conference, Dallas, TX, 3
- 6 March 1998.
14. Jones, S.C. and Yaeger, L.G., "Automated Mud System Enhances
Well Control". Petroleum Engineer: September 1974.
15. Kennedy, J.L., "Computer drilling system can provide
optimization, rig control". Oil and Gas Journal, 69(18):10 May
1971, 61 - 64.

16. Killalea, M. et al., "Rigs of the Future, Wells of the Future".

Drilling Contractor, 63(1): January/February 2007, 16 - 64.
17. Naegle, P.N., Dant, R.E., Dieball, K.J., Stephenson, S.V. and
Padgett, P.O., 'Cement Mixing and Pumping System and Method
for Oil/Gas Well", 5,289,877, 1 March 1994.
18. Parsons, C.P., "Prospecting with rotary equipment". Oil and Gas
Journal: 7 June 1928.
19. Rach, N., M., "New rig control system provides closed-loop
drilling automation". Oil and Gas Journal, 105(27): 16 July 2007.
20. von Flatern, R., "Automating the drill floor". Offshore Engineer,
28(6): 10 June 2003.
21. Womer, K., "Use of 21st Century Computer and Communications
Technologies to Make Effective Drilling Decisions" SPE 101516DL, presented at SPE Distinguished Lecture Series,
22. Womer, K., Koederitz, W.I. and Guggart, M.I., 'Automated Rig
Control Management System", 6,944,547, 13 September 2005.
23. Zinkgraf, H.L. and Hammett, D.S., "Drilling Improvements
Using Power Swivels" IADC/SPE 11403, presented at IADC/SPE
1983 Drilling Conference, New Orleans, LA, 20 - 23 February
1983.