From High Technology to Solutions: The Experience of iXSea
Thierry Gaiffe
President, IXSEA
55, Avenue Auguste Renoir,
78160 Marly-le-Roi
France
Tel: + 33 1 30 08 98 88
Fax: + 33 1 30 08 88 01
info@ixsea.com
is of high technical level.
1 Introduction It is with this pool of men and women passionately com-
mitted to science and engineering that iXSea is inventing
the technologies and concepts the sea will require tomor-
In any business school in the world, you will learn that row.
there are just two ways to develop and lead a company: a
manager must choose between being “techno-driven” or To illustrate the above, I propose to describe in detail one
“market-driven”. However, all previous success stories technology invented in our laboratories which have each
demonstrate that the reality is far removed from this highly yielded products that have made possible the emergence of
theoretical vision: in most cases, successful development systems and tools used today by oceanographers and sur-
results from an overlap between an idea and a need. This veyors the world over: the FOG, or Fiber Optic Gyroscope.
means that in practice the main issue for a manager is to Our history in the field of FOG technology
[1]
goes back
invent new concepts while getting a feel for market orienta- nearly 20 years, and we have provided key components that
tion and trying to reach the best decision as to ‘when and have enabled this technology to emerge at industrial level
how’ the company can bridge the two. for the most stringent applications in terms of performance
[2]
In 2000, IXSea focused on a single goal: ensuring that it and reliability .
was capable of providing the right technical solutions ocean
exploration would be looking for throughout the coming A FOG is a gyrometer, that is to say it is a sensor that can
century. measure instantaneously the rotational speed of a mobile
platform.
Why this choice?
The FOG is based around the Sagnac Effect discovered in
There can be no doubt that one of the challenges of the 21st [3]
century will be to become the masters of the largest part of 1917 by a physicist called Georges Sagnac . This is a
our planet – its oceans – and it is also obvious that at pre- relativistic phenomenon not dissimilar to the familiar Dop-
sent we have no idea of the means we will employ to ac- pler effect.
complish that. However, what is absolutely certain is that To understand it properly, we need to look at the diagram
all those involved in this challenge, scientists, oil and gas illustrating the basic principle: a ‘ring’ interferometer, that
industries, naval forces, etc… have their eyes on the same is to say an interferometer composed of a coil of optical
goals: deeper, faster and easier! fiber which loops back on itself (cf. figure 1).
For example, in just five years iXSea brought fiber optic
gyro technology to unmanned underwater vehicles, de-
signed USBL plug-and-play systems, democratized Syn-
thetic Aperture Sonar and invented real-time magnetic
imaging. We saw here a need for interdependence between
wide-ranging R&D in optoelectronics, acoustics, signal
processing and NMR, of which we had complete mastery in
our research labs, and easy-to-use solutions for the most
demanding applications.
The main purpose of the present paper is to look again at
some of our technologies, and see how they could solve
some of the most challenging underwater problems of the
early 2000s, and also to describe the solutions on which Figure 1: The basic principle underlying the FOG
iXSea is working in order to meet the latest challenges in
positioning and imagery.
The light from a single optical source is divided into two
2 The technology upstream parts by a beam splitter, and the two light waves thus gen-
erated travel around the optical fiber coil in opposite direc-
We see ourselves as an enterprise made up of scientists and tions. When they emerge from the coil, they have on the
engineers. At every level in our managerial structure, iXSea face of it traveled over the same optical path and they
breathes the same air of high technology and determination should therefore be in phase.
to push back further and further the known technical fron-
tiers, the starting point being our human resources: the What the Sagnac Effect in fact says is that when the coil is
educational background of three-quarters of our workforce made to rotate the two waves, when they emerge, will be
1
out of phase (delayed) to a degree proportional to the veloc- on.
ity of the rotation.
Like all gyrometers, FOGs are essentially the main compo-
nents of Inertial Navigation Systems (INS), units which
contain a minimum of three gyrometers and three acceler-
ometers.
At every moment an INS can measure the position and
The coefficient of this proportionality will depend on the attitude of a moving solid platform, which is generally a
geometric sizing of the coil (length L and diameter D), and vehicle of some kind, for which we want to determine in
especially its apparent surface area: the phase shift caused real time six navigation parameters: latitude, longitude,
by the Sagnac Effect is in fact a measure of the flux of the altitude, heading, roll and pitch.
rotational vector through the optical fiber coil, a little like The accuracy of an INS in fact derives from the accuracy of
magnetism or the familiar phenomenon of induction which its gyrometers, and this will drift with the passage of time.
describes how a current can be created by the flux of a This might on the face of it seem paradoxical, since intui-
magnetic field through an electrically conductive coil. tively you could consider that if the position is defined by
It is easy to understand why FOG technology enables cur- the double integration of the accelerometers, the position
rently unequalled levels of performance to be achieved: by would drift according to the square of time and the stability
increasing the apparent surface area of the coil, that is to of the accelerometers. But this is not the case!
say its diameter and its length, it is possible to increase the What is often forgotten is that navigating along a straight
proportionality coefficient and therefore the detection line on the surface of the Earth, assumed to be roughly
sensitivity. spherical, actually amounts to traveling on a great circle,
The diagram provided above simply illustrates the underly- and therefore rotating. This means in fact that the stability
ing principle and no FOG could actually operate in this way of a gyrometer will determine, taking the Earth’s radius
for a number of reasons, and in particular the dimensional into account, the drift in the position given by the naviga-
stability of the optical alignments at the interfaces, and the tion unit [6].
fact that the sensor has zero sensitivity at low rotational
speeds, since the intensity of the interference varies with
the cosine of the phase shift. The optimum configuration INS Drift in Nm/h = 60 x Gyro Bias in deg/h
for a FOG offering best performance is shown in figure 2.
Class 0.01 deg/h gyrometers therefore give us INS units
Coil
whose stability in ‘pure inertial’ mode will be 0.6 Nm/h,
Integrated Optical
coupler
COUPLEUR
Circuit (IOC)
τ
which is quite sufficient for aircraft navigation, but far from
SOURCE 2X2 good enough for an AUV!
To arrive at satisfactory performance it is necessary to help
the INS out using other types of sensor made statistically
D/A
Modulation + Ramp Modulation + Ramp
VERS DETECTEUR for high Ω value for smaller Ω value
A/D
ET
CARTE TRAITEMENT independent of the navigation unit by optimized real-time
±1
filtering of Kalman type. Where AUVs are concerned, we
can use auxiliary acoustic systems, usually Doppler logs
(DVLs), or in some cases acoustic positioning systems
Figure 2: Optimum FOG configuration (USBL) [7].
Given the high accuracy of a DVL’s stable measurement of
the speed of travel over the sea bottom, of the order of
By using this type of configuration, known as an ‘all-digital millimeters per second, the speed drift of the navigation
closed-loop hybrid’, the performance obtained makes it unit can be corrected using this information, which will
possible to attain the Holy Grail of rotation detection: improve, thanks to the Kalman filter, the information on
stability better than 0.01 degree per hour, which in turn gyrometer and accelerometer bias, which will also by the
enables what is called inertial navigation, that is to say its same token enhance the accuracy of the determination of
use as a ‘black box’ without a GPS. In actual fact, the position and attitude. This means that the Kalman filter
FOGs produced by iXSea perform even better than this and method for correcting the navigation unit is infinitely more
continue to improve as time goes by: for example, we effective than simple dead reckoning. In the last analysis,
manufacture the gyrometers for European Earth Observa- the accuracy of the positioning will be limited only by the
tion satellites [4] and autonomous navigation units for DVL’s own drift, which is around 3 m/h.
nuclear attack submarines, and such gyrometers are in the
0.001 deg/h class, or 10-4 of the earth’s rotation [5]. After having placed on the market the first optical fiber
gyrocompass in 1997 – the Octans™ – iXSea successfully
introduced the first optical fiber INS in 2001 – the Phins™
3 From Technology to Equipment
– which is also the only INS specifically developed for use
Our mastery of the technology required upstream gives us in AUV navigation.
independence and the capacity to adapt to potential market
Its dual characteristics led to its adoption by most builders
requirements.
of AUVs. First and foremost, it is a specific feature of FOG
Continuing our demonstration in relation to the FOG, our technology that compared with the older gyrolaser technol-
capability for high performance made it possible for us to ogy (RLG) it consumes much less power and offers signifi-
propose innovative solutions to our customers very early cantly greater reliability and longevity, qualities the space
2
industry found attractive. Secondly, the Kalman filter in the specific to a given application.
Phins was developed on the basis of the specific character-
istics of the marine environment and of acoustic sensors, For example, it is standard practice today to correct an
which is not the case for more conventional INS units AUV’s INS using data from a USBL which is itself cor-
developed for civil or military avionics use. rected by a GPS. We have also carried out development in
the reverse direction: when the USBL loses the GPS data, it
The inertial navigation units made by iXSea are not used is the inertial unit that corrects the USBL at the surface,
only for autonomous vehicle navigation, but also for stabi- thereby enabling the surface platform to retain its naviga-
lizing platforms and compensating for swell-induced tion capability.
movement in supporting structures.
In other respects, the merging of acoustic and inertial tech-
On the basis of IXSea’s many years of experience in acous- nologies commenced by iXSea just a few years ago has not
tic positioning, and particularly its development and manu- focused solely on positioning issues.
facture of the world’s longest-range USBL, the Posido-
nia™, we came up in 2003 with the idea of merging USBL Indeed, we decided two years ago to develop a Synthetic
and INS to create a new type of totally portable USBL that Aperture Sonar (SAS), capitalizing on our mastery of iner-
can be easily deployed on any type of ship without prior tial technologies, acoustic transducers and real-time algo-
calibration. This product was the GAPS™ [8]. rithms. We placed our own SAS on the market recently, in
March 2006: this is the Shadows™.
Figure 4: Shadows
The basic principle underlying the SAS is to use the speed
of the structure on which it is installed to create syntheti-
Figure 3: OCTANS, PHINS and GAPS cally a virtual antenna larger than the physical antenna,
thereby enhancing the resolution that would otherwise be
obtained conventionally with a standard antenna of the
Specifically, the fact that the GAPS includes an attitude
same length as the physical antenna.
detection unit accurate to 0.01 deg which requires no cor-
rections for attitude offset, in addition to its use of the most The idea is to reconstitute along the path of the antenna its
modern chirped acoustic technology, enables the position displacement during the system’s coherence time.
of objects to be determined up to distances of 4,000 meters
with a CEP (Central Equal Probability) accuracy of 0.2% It is of course possible to use the redundancy of the sonar
slant range. image to reconstitute the antenna’s curved trajectory, but
although the algorithm is effective on paper, in actual prac-
The 3-dimensional design of the acoustic antenna has been tice if the design of the SAS is based on correction of the
specially engineered to make it possible to position objects trajectory by the image it will be fairly unstable and sensi-
within a 200-degree cone, enabling ROVs to be tracked tive to outside disturbance.
vertically at great depths at the same time as towed fish
behind the ship traveling horizontally and just at the sur- The idea here is to use an INS for accurate measurement of
face. the antenna’s micro-displacements to reconstitute the
curved trajectory of the physical antenna in order to synthe-
But one of the main advantages of the GAPS is its proprie- size the virtual antenna.
tary algorithm developed by iXSea’s ablest mathemati-
cians. The Kalman filter optimizes the merging of data In practice, it is our detailed knowledge of the behavior of
from the inertial navigation unit and the acoustic antenna our INS and the very high quality of our FOGs that made it
and calculates in real time the best estimate of the trajectory possible to rise to this challenge. Once again, in the last
of the submarine vehicle. Among other things, this makes it analysis what is important is to get the technology right …
possible to calculate the position of an object at all times,
without delay and at an arbitrarily chosen frequency, mak-
ing the monitoring of a remotely operated object such as a References
ROV more straightforward and intuitive.
[1] H. Lefevre, “The Fiber Optic Gyroscope”, 1993,
Artech House.
4 From the Equipment to the Solution
[2] T. Gaiffe, K. Wandner, Y. Cottreau, N. Faussot, P.
Inertial, acoustic and hybrid products for navigation and
Simonpietri, H. Lefevre, “Low Noise Fiber Optic
positioning can also be mutually interfaced and supplement Gyroscope for the Sofia Project”, 1999, Sympo-
each other to provide a solution that will match the needs sium Gyro Technology (DGON 99), Stuttgart.
3
[3] G. Sagnac, “L’éther lumineux démontré par l’effet
du vent relatif d’éther dans un interféromètre en ro-
tation uniforme”, [The existence of light ether
demonstrated by the effect of the relative ether
wind on a uniformly rotating interferometer],
Comptes Rendus de l’Académie des Sciences Vol.
95, 1913, pp 708-710.
[4] E. Willemenot et al, “Very High Performance FOG
for space use”, Symposium Gyro Technology,
Stuttgart 2002
[5] Y. Paturel, V. Rumoroso, A. Chapelon, J. Hontaas
“Marins, the New High Performance Navigation
System for Submarines”, UDT 2005, Amsterdam.
[6] T. Gaiffe, “ Phins: an Inertial Navigation System
developed specifically for AUV Control and Navi-
gation”, Underwater Intervention 2001, New Or-
leans.
[7] T. Gaiffe, “Global Underwater Positioning and
Navigation Solutions”, ON&T, November 2003,
pp 55.
[8] T. Gaiffe, “GAPS, the USBL that combines acous-
tic and inertial technologies”, Ocean Systems 2003.
