Tunnelling and

Underground Space
Technology
Tunnelling and Underground Space Technology 19 (2004) 597605

incorporating Trenchless
Technology Research

www.elsevier.com/locate/tust

Modelling TBM performance with articial neural networks

A.G. Benardos *, D.C. Kaliampakos
School of Mining and Metallurgical Engineering, National Technical University of Athens, GR 15780 Athens, Greece
Received 29 September 2003; received in revised form 30 January 2004; accepted 11 February 2004
Available online 18 March 2004

Abstract
Assessing TBM performance is an important parameter for the successful accomplishment of a tunnelling project. This paper
presents an attempt to model the advance rate of tunnelling with respect to the geological and geotechnical site conditions. The
model developed for this particular task is implemented through the use of an articial neural network (ANN) that allows
the identication and understanding of both the way and the extent that the involved parameters aect the tunnelling process. The
model described in the paper is customised for the construction of an interstation section of the Athens metro tunnels, where the
ANN generalisations provided precise estimations regarding the anticipated advance rate.

2004 Elsevier Ltd. All rights reserved.
Keywords: TBM tunnelling; Articial neural networks; Advance rate modelling

1. Introduction
The performance analysis of tunnel boring machines
(TBM) and the development of accurate assessment
models have been, and still are, the ultimate goals of
many researchers (Tarkoy, 1973; McFeat-Smith and
Tarkoy, 1979; Bruland et al., 1988; Bruland, 1999; Sharp
and Ozdemir, 1991; Nelson, 1993; Barton, 2000), as the
reliable estimation of the excavation rate is proved to be
crucial for the projects economics (Alber, 2000). In this
pursue of developing the most consistent model, the
paths followed have been numerous (Lislerud, 1988;
Laughton and Nelson, 1996; Blindheim et al., 2002). This
derives from the fact that a variety of input parameters
and methodological approaches have been used.
Beyond mathematical formulae and analytical solutions, methods utilising articial intelligence have not
been introduced until recently (Bruines, 1988; Alvarez
Grima et al., 2000; Okubo et al., 2003). Nevertheless,
they showed very promising results, demonstrating their
strong potential in coping with this particular issue. In
the majority of these research eorts, the main objective
*

Corresponding author. Tel.: +30-210-7722182; fax: +30-2107722156.

E-mail address: abenardos@metal.ntua.gr (A.G. Benardos).
0886-7798/$ - see front matter
2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tust.2004.02.128

is to model the tunnelling process and make the performance assessment, based on the experience gained
and the data gathered from past projects. However, even
though probing risk conditions and identifying vulnerable areas that may disrupt the work progress have been
incorporated in the models of many researchers (Einstein et al., 1992; Sineld and Einstein, 1996), they have
not yet been fully addressed, leaving room for further
research. These problems are more intense in tunnelling
projects constructed in complex geological formations
(Barla and Pelizza, 2000) and especially in urban areas
where the low construction depth and the external
loading from the buildings increase risk conditions
(Duddeck, 1996; Eisenstein, 1999).
This paper deals with the modelling of the TBM
performance emphasising on the identication of the
performance oscillations throughout the tunnelling period. This is made possible by the development of an
articial neural network (ANN) capable of learning
from the tunnelling experience and generalise solutions
making prognosis for new input data. Hence, the main
aim is to produce a tailor-made model, utilised during
the construction period, capable of providing estimates
of the expected tunnelling advance rate. The model can
also be used in another context; to assist in the identication of dicult ground conditions that may disrupt

598

A.G. Benardos, D.C. Kaliampakos / Tunnelling and Underground Space Technology 19 (2004) 597605

the work cycle and impair the TBMs performance.

Both outputs, modelling of the TBMs advance rate and
identication of the risk-prone areas, are essential
knowledge for the engineers in order to ensure the nancial and scheduling credibility of the tunnelling
project.
In the following paragraphs an overview of the
ANNs is given, along with the presentation of the
methodological approach, in a more detailed manner.
Furthermore, a case study of the proposed methodology
is introduced, using an interstation section of the Athens
Metro tunnels as an illustrative example.

2. Articial neural networks

The development of ANN started as an attempt to
understand the operation of the human brain and mimic
its assessment capabilities. In other words, to be able to
decide and act under uncertainty or even deal with situations having limited previous experience. ANNs are
mathematic models consisting of interconnected processing nodes (neurons) under a pre-specied topology
(layers).
Neural networks have a strong similarity to the biological brain and therefore a great deal of their terminology is borrowed from neuroscience. Their basic
characteristic is the ability to perform massively parallel
computing of the input stimulus (data), contrary to the
custom mathematic models that are based rather on a
serial process of mathematical and logical functions
(Fausett, 1994). Another advantage of the ANNs is their
exibility in data processing, as no deterministic mathematical relationship of the examined components is
required. Instead, once the data is introduced, in a
causeeect mode, the network identies the existing
relationships, learns and mimics their behaviour by adjusting the strength of the links between the neurons
(connection weights). Thus, they cannot be programmed
but they are rather taught through case experience. As a
result, soon after the ANNs training, given an existing
dataset, estimates can be drawn for another specic data
input. Thus, the trained network can generalise and give
estimates for uncertain conditions or even incomplete
data (Sietsma and Dow, 1991). The main disadvantage
of ANNs is that an explicit determination of the parameters weighting is not an easy task or it may not
even be possible in large and complex network architectures (Menhrotra et al., 1997).
The ANN operation is based on the following:
Data processing occurs in a number of simple processing units (neurons), which have signal inputs
and outputs.
The neurons bonding is made through connection
links, each one of them having a corresponding
weight that multiplies the signal.

Fig. 1. Illustration of an articial neural network structure (after

Fausett, 1994).

Each neuron applies an activation function to the signal input to control the signal output.
In general, a typical ANN topology is consisted by a
set of layers; the input layer, one or more hidden layers
and the output layer. Each one of them includes a
certain number of neurons, specied by the ANN architecture. Accordingly, each neuron is linked to neighbours with varying coecients of connectivity that
represent the weighting of these connections. The topology of a simplied ANN is presented in Fig. 1.
In this simple model, there is one hidden layer having
only one neuron. Each neuron of the hidden layer(s) is
interconnected to all others found in the input and
output layers. The hidden layers are the most important
element of the network as this is the particular part
where the network learns the interdependencies of the
model. This learning procedure is accomplished by adjusting the connection weights, impelling the overall
network to generate the matching results. In this manner, changing the connection weights (training) causes
the network to learn the solution for a given problem.
In the topology of Fig. 1, each neuron of the input
layer (X1 ; X2 ; X3 ), sends out its weighted signal to the Y
neuron found in the hidden layer. The combined input
signal in the Y neuron has the following form:
Yin w1
x1 w2
x2 w3
x3 ;
where, xi is the signal of the ith input neuron, wi the
weighting factor of the ith neuron.
The input signal (Yin ) is introduced to the activation
function of the Y neuron and signalled to the neurons of
the output layer, Z1 ; Z2 following the general form:
y f Yin taking into account the weighting of the
connection links, namely, v1 and v2 .
The type of ANN used in this paper are the feedforward neural networks, which are the most widely
used. They are commonly applied to problems where a
set of input vectors should be corresponded to another
specied set of output vectors.
The training procedure consists of a sequential data
feed into the network, followed by the comparative
evaluation of the corresponding output provided by the
ANN and the actual result. The network adjusts the

A.G. Benardos, D.C. Kaliampakos / Tunnelling and Underground Space Technology 19 (2004) 597605
Adjustment of connection
weighting

Data input

Comparison
with actual
data

Hidden Layers

Fig. 2. Training procedure of a feed-forward ANN with two hidden

layers.

weighting of the connection links in a continuous eort

to produce the results that would best correspond to the
training dataset. A complete pass of all the input data
through the network consists a training epoch and
usually a great number of epochs is required for the
residual error to converge below a pre-specied threshold. A schematic illustration of a feed-forward ANN
training is given in Fig. 2.
The goal is to train the network so as to achieve a
balance between its capability to memorise the training
input vectors and its capability to generalise, i.e. to
produce outputs for input vectors that are similar but
not identical to the ones used for the training. In other
words, the ANN should avoid data overtting phenomena, but should have the ability to produce a consistent generalisation.

3. Model synthesis
The model concentrates on the tunnel construction
period in soft ground environments. The whole idea
follows the ANN philosophy, that is, to analyse the
experience gained from the tunnel boring process and to
correspond it to a set of selected data. This causeeect
request is used in the ANN so as to identify the interactions between the data and to come up with the exact
weighting of the parameters involved, which will nally
determine the generalisation accuracy.
The models inputs are based on data relating to the
geological and geotechnical characteristics of the subsurface and the specic site conditions. Although machine characteristics (e.g. thrust, torque) are very
important for the overall TBM performance, in the case
where tunnelling is performed in soft rock or complex
ground formations, the properties of the ground medium tend to be the most inuential ones, as they govern
the type and extend of possible failures. Subsequently,
encountering ground conditions dierent from the

599

TBMs working envelope, aect the achieved tunnelling

rate (Deere, 1981) and can give rise to claims (Buchi,
1998). Thus, the model considers the geological setting
to be the most dominant factor for the TBM performance, as many researchers have also noted (Tarkoy,
1981; Nelson, 1993; Sapigni et al., 2002).
In this way, it is assumed that the characteristics of
machine operation remain unchanged and all possible
problems and downtime are a direct eect of the geotechnical conditions. Even though downtime is also inicted by machine failures, logistics support problems,
etc., the real question is how the TBM performance is
aected by ground conditions and the aforementioned
assumption is made exactly so as to be able to evaluate
this particular issue. Having that in mind, the data
gathering procedure concentrates on obtaining information about the subsurface conditions encountered
and the scheduling data, yet excluding all machinery
occurred failures (e.g. power failures, belt replacements,
etc.).
The selection of the parameters used in the model was
made having in mind their capability to credibly represent the ground behaviour, hydrogeological environment and site-specic conditions (Benardos, 2002).
These parameters are easily collected in the site-investigation phase and are available to all design stages of the
project, without the need for implementing special investigation techniques. More specically, these parameters, are
rock mass fracture degree as represented by RQD
(P1 ),
weathering degree of the rock mass (P2 ),
overload factorstability factor (N) (P3 ),
rock mass quality represented by RMR classication
(P4 ),
uniaxial compressive strength of the rock (UCS) (P5 ),
overburden-construction depth (P6 ),
hydrogeological conditions represented by the watertable surface relative to the tunnel depth (P7 ),
rock mass permeability (P8 ).
Many of them have already been proposed as indicators of the tunnelling eciency. For example, the
fracture degree of rock masses is extremely important to
TBM tunnelling (Deere and Deere, 1988), the overload
factor, rst introduced by Peck (1969), can provide information about the face stability conditions. In addition, the compressive strength is inuencing TBM
performance, while RMR is very important as it denotes
the tunnels stand-up time and is also used in TBM
performance analysis (Sapigni et al., 2002). Finally, as
Terzaghi (1950) noted the hydrogeological conditions
and the presence of water is directly or indirectly linked
to the problems occurring in soft ground tunnelling.
From a rst point of view, it appears that there is a
degree of interconnection between the parameters (e.g.
RDQ is incorporated in RMR), that might bias the

600

A.G. Benardos, D.C. Kaliampakos / Tunnelling and Underground Space Technology 19 (2004) 597605

nal results. Nevertheless, each parameter describes a

specic issue and the information provided can further
assist in the deeper understanding and clarication of
the possible problems or mishaps encountered. For
instance, even when there are cases with the same
overall RMR values, where similar behaviour is expected, the dierentiation in the fractures and joint
network, consequently and in the RQD values, might
lead to dierent stability conditions; stable conditions
in one case, while experiencing instabilities and collapses on the other.
The case study used for the model development is an
interstation tunnel of the Athens Metro. The geological
setting is a system of low-level metamorphic sedimentary weak rock consisted of interbedded marly lime-

stones, calcareous sandstones, siltstones, conglomerates,

phyllites and schists. The formations are intensely
thrusted, folded and faulted with a variable and erratic
degree of weathering and alteration (Kavvadas et al.,
1996). The examined tunnel is located between the
Katehaki and Panormou stations (Fig. 3). It is the longest interstation tunnel in the Athens Metro, until now,
having a total length of 1129.36 m (Attiko Metro SA,
1995a). The examined tunnel length is approximately
1077 m, excluding the rst 53 m (learning curve period).
The area is divided in 11 control areas (segments), in
which the data is collected and the assessment of the
selected geological properties is made (Fig. 3). All data
from boreholes have been spatially modelled so as to
identify the properties especially within the 12 m thick

Fig. 3. Layout of the examined Athens Metro tunnel.

Fig. 4. Spatial modelling of the RQD values in the area of the examined tunnel.

A.G. Benardos, D.C. Kaliampakos / Tunnelling and Underground Space Technology 19 (2004) 597605

For each segment, a corresponding value for every

principal parameter is taken. Allocating a representative
value for the parameters is accomplished by the spatial
modelling of the parameters value and by the incorporation of statistical distribution that mimics the parameters behaviour in each segment (Benardos, 2002).
In Fig. 4 the spatial modelling of the RQD values is illustrated, for the whole analysis area.
In the next step, the data is categorised in four interval scale classes, from 0 to 3, where 0 denotes the
worst case and 3 the best. The limits taken in every class
are representative of the specic site conditions and the
machine characteristics. In the case of the Athens metro,
the tunnel is constructed in relative low depth and, in
general, in weak rock conditions with a double shield
TBM machine. The rating of each parameter is presented in Table 1.
The limits of the proposed rating transforms the
continuous data to a discrete probability structure, a
form that is nally used as input to the model. More
specically, the data is introduced to the ANN as the
expected values (EV) of the parameters (Table 2). For
example, given V1 ; V2 ; . . . ; Vn values having a respective
probability of occurrence P1 ; P2 ; . . . ; Pn , the expected
value of the variable X , is estimated as
n
n
X
X
EX  EV
Pi
Vi ; while;
Pi 1:

Table 1
Rating of the principal parameters
Value class

Rating

Rock mass fracture degree RQD

<10
1030
3060
>60

0
1
2
3

Overload factor (N)

>5
35
1.253
<1.25

0
1
2
3

UCS (MPa)
<2
215
1540
>40

0
1
2
3

Water table surface (m)

>10
510
05
<0

0
1
2
3

Rock mass weathering

Compl. Weath. CW
High Weath. HW
Med. Weath. MW
SW, fresh

0
1
2
3

Rock mass rating RMR

<10
1030
3060
>60

0
1
2
3

601

i1

i1

Table 3
Tunnelling advance rate data in each one of the control segments

Overburden (m)
<7.5
7.512.5
12.517.5
>17.5

0
1
2
3

Permeability (m/s)
<104
104 106
106 108
>108

0
1
2
3

stratum that the tunnel is actually being built in, ranging, along the chainage, from the level of +120 m to the
level of +156 m.

Segment

Average AR
(m/day)

Max AR
(m/day)

Min AR
(m/day)

1
2
3
4
5
6
7
8
9
10
11

4.00
4.54
6.25
4.35
9.82
9.09
16.67
11.11
10.85
12.50
14.07

8.8
8.8
10.4
13.5
12.1
13.7
21.0
18.3
17.0
17.3
14.8

0.0
0.0
2.8
0.0
0.5
7.3
14.7
4.4
6.1
1.6
10.4

Table 2
Expected values of the principal parameters in each segment
Parameter

Seg1

Seg2

Seg3

Seg4

Seg5

Seg6

Seg7

Seg8

Seg9

Seg10

Seg11

Rock mass fracture degree

Rock mass weathering
Overload factor
Rock mass rating
UCS
Overburden
Water table surface
Permeability

0.13
2.52
1.07
0.00
0.48
0.42
3.00
1.92

0.88
2.52
0.89
0.00
0.57
1.00
2.32
1.97

0.90
2.24
1.92
0.36
0.97
1.17
1.71
1.95

0.64
1.97
1.99
0.93
1.06
1.97
1.00
1.89

0.72
1.99
2.73
1.10
1.68
2.86
0.23
1.86

1.37
1.89
2.16
1.83
1.31
2.35
0.02
1.69

1.62
1.95
2.49
2.00
1.28
1.16
0.94
1.90

1.26
1.93
2.28
1.49
1.20
1.45
1.40
1.82

0.55
1.96
2.43
1.08
1.16
1.13
2.17
1.86

0.66
1.94
2.61
1.00
1.21
0.99
2.40
1.76

0.74
1.93
2.43
1.00
1.15
0.88
2.75
1.81

602

A.G. Benardos, D.C. Kaliampakos / Tunnelling and Underground Space Technology 19 (2004) 597605

The tunnelling advance rate, achieved in each segment, is also introduced into the ANN model. Hence,
the input vector of the principal parameters is tallied to
the output vector of the mean achieved advance rate, in
each segment (Table 3), expressed in m/day (Attiko
Metro SA, 1995b). Note that all external origin delays
(e.g. strikes, maintenance, etc.) have not been taken into
account.

The LevenbergMarquardt algorithm, selected for

training the ANNs, is a variation of the classic backpropagation algorithm that, unlike other variations that
use heuristics, relies on numerical optimisation techniques to minimise and accelerate the required calculations, resulting in much faster training (Demuth and
Beale, 1994). More specically, the direction in which
the search is made is described by the following equation:
xk1 xk  A1
k
gk ;

4. ANN development
In order to proceed with the development of the
ANN model, the dataset of the whole 11 analysis
segments has been divided into two subsets. The rst
one (training subset A) is used for the ANNs training,
whereas the second (test subset B) is used for the
validation of the models generalisation capability.
Special focus is given on the second subset (B), as the
network consistency should be ensured for the whole
spectrum of cases. Thus, a set incorporating the most
representative cases, in terms of the achieved advance
rate, has been selected. Apparently, segments no. 2, no.
7 and no. 9, are selected as they represent the worst, the
best and an average case. Consequently, the two subsets
are comprised by the data collected in the following
segments: A {1, 3, 4, 5, 6, 8, 10, 11} and B {2, 7, 9}.
The neural network toolbox of the Matlab software
package has been used for building the ANN code and
performing the training and testing of the model.

where, Ak is the Hessian matrix of the error function at

the current values of weights and biases and gk is the
gradient of the error function.
Since the error function has the form of a sum of
squares the Hessian matrix can be approximated as
A JT
J;
and the gradient as
g J T
e;
where, J is the Jacobian matrix, which contains rst
derivatives of the network errors with respect to the
weights and biases, and e is a vector of network errors.
Finally, the search direction is given by
1

xk1 xk  J T
J l
I

J T
e:

In the case where the scalar l is zero, this is just

Newtons method, using the approximate Hessian matrix. When l is large, this becomes gradient descent with
a small step size. Newtons method is faster and more
accurate near an error minimum, so the aim is to shift

Fig. 5. ANN code development in the Matlab Editor/Debugger environment.

A.G. Benardos, D.C. Kaliampakos / Tunnelling and Underground Space Technology 19 (2004) 597605

603

Fig. 6. Training error vs training epochs.

towards Newtons method as quickly as possible. Thus,

l is decreased after each successful step (reduction in
performance function) and is increased only when a
tentative step would increase the performance function.
In this way, the performance function will always be
reduced, at each iteration, of the algorithm.
The ANNs performance is assessed in terms of the
relative error level (D) achieved, between the actual and
the predicted advance rate (AR), following the expression:
D

ARactual  ARpredicted
:
ARactual

This criterion can provide a clear aspect regarding the

ANN behaviour and moreover makes possible the
comparison between the ANN results and other methods or theoretical models focusing on advance rate
prediction.
A number of test runs have been conducted in order
to come up with the network architecture that produces
the more consistent results. From the various network
architectures that were examined, two particular ANN
architectures (8  9  4  1 and 8  10  7  1) proved
to be more promising as they responded quite well to the
training process. The ANN that was nally selected
followed the rst architecture, namely the 8  9  4  1
form. This particular structure type means that the
ANN has a total of 4 layers, with 8 neurons in the input
level, same as the number of the parameters, two hidden
layers with 9 and 4 neurons respectively, followed by 1
neuron in the output layer that eventually generates the
value of the advance rate. The code used for the ANN

development is presented in Fig. 5. The mean squared

error (MSE) of training for this particular ANN model
approximates 1.4  1027 and is attained after 103
training epochs, as illustrated in Fig. 6.
The results generated from the trained model were
very satisfactory (Table 4). The relative error between
the model outputs and the validation subset ranges in
the region of 6% and 8%, reaching a maximum of about
8.4%, a level that is quite acceptable. Furthermore, the
ANN behaviour shows that the results are consistent in
all the validation subset segments, element of major
importance for an accurate and eective measurement of
the TBM performance.
In Fig. 7 a surface plot of the model is presented. It
has been constructed in relation with the RMR and
UCS parameters for a given RQD value of 0.5; the
values of all other parameters are taken equal to their
mean values. This nomograph can be used as a way of
presenting the eect of the selected parameters on the
TBM advance rate.

Table 4
Comparison between the ANN generalisation output and the actual
advance rate data
Segment

ANN generalization
results

Actual data

Relative error

2
7
9

4.8545
17.6875
9.9424

4.54
16.67
10.85

0.0693
0.061
)0.0837

604

A.G. Benardos, D.C. Kaliampakos / Tunnelling and Underground Space Technology 19 (2004) 597605

consistent results. Thus, the model can nd its optimal

use in cases of intense urban underground development
(e.g. subways, sewage tunnels, etc.) as the operations are
taking place in roughly the same geological setting, with
the same methods and tools, extensive data is available
from past projects and a constant data ow can be expected from the worksites.

References

Fig. 7. Surface plot of the expected advance rate with respect to RMR
and UCS for a given RQD value.

5. Conclusions
The development of articial intelligence methods for
modelling TBM performance has been well accepted
through the scientic community, as the various attempts made in that eld proved their eciency. The
ANN system used in this paper demonstrated very satisfactory results in predicting the achieved advance rate
for the case study in question. The resulting remarks can
be drawn hereinafter:
Once trained, the ANN can become a practical othe-self tool for the prediction of the tunnelling advance rate. Its ease of use and its straightforwardness
in giving the results can allow its utilisation even for
on-site assessments.
The open source code increases the models exibility,
allowing also the insertion of additional data enhancing the prediction accuracy of the nal results, even
on daily basis.
The prediction of the TBM advance rate can be used
for the identication of risk-prone areas. As the model is based on geotechnical data, a drop in the advance rate indicates that the area in question may
eventually pose threats to the tunnelling process
and special attention should be paid.
The ANN model can also be utilised for a projects
strategic development. Thus, it can be used either
for choosing the best tunnel alignment from a number of alternatives, or selecting the most appropriate
ground improvement technique if needed to overcome any diculties or major downtime due to adverse ground conditions. In both cases, scenario
analysis can be performed by changing the values of
the input parameters, with respect to the proposed
tunnel alignment or technique followed. Thus, a direct comparison can be made in nancial terms, regarding the best possible selection that would
ensure the projects success.
Finally, it should be noted that in all cases a number
of records should be available in order to come up with

Alber, M., 2000. Advance rates of hard rock TBMs and their eect on
project Economics. Tunnelling and Underground Space Technology 15 (1), 5564.
Alvarez Grima, M., Bruines, P.A., Verhoef, P.N.W, 2000. Modelling
tunnel boring machine performance by Neuro-Fuzzy methods.
Tunnelling and Underground Space Technology 15 (3), 259269.
Attiko Metro SA, 1995a. Interstation Katehaki Panormou: Geological Geotechnical Data, Athens.
Attiko Metro SA, 1995b. Interstation Katehaki Panormou: General
Construction Report, Athens.
Barla, G., Pelizza, S., 2000. TBM tunnelling in dicult ground
conditions. GeoEng 2000, Melbourne, 2000.
Barton, N.R., 2000. TBM tunnelling in jointed and faulted rock,
Balkema.
Benardos, A., 2002. Hazard identication in the construction of
underground excavations using tunnelling boring machines (TBM)
The case of the Athens Metro. Ph.D. Thesis. National Technical
University of Athens.
Blindheim, O.T., Grv, E., Nilsen, B., 2002. The Eect of Mixed Face
Conditions (MFC) on Hard Rock TBM Performance. ITA
Assembly, Sydney.
Buchi, E., 1998. TBM tunnelling contracts with increased potential for
claims? In: Franzen, T. (Ed.), International Conference on Underground Construction in Modern Infrastructure, pp. 213218.
Bruines, P., 1988. Neuro-fuzzy modelling of TBM performance with
emphasis on the penetration rate. Memoirs of the Centre of
Engineering Geology, Delft, no 173.
Bruland, A., 1999. Prediction model for performance and costs.
Norwegian TBM Tunnelling, Norwegian Tunnelling Society, pp.
2934.
Bruland, A., Johannessen, B.E., Lislerud, A., Movinkel, T., Myrvold,
K., Johannessen, O., 1988. Hard rock tunnel boring. Project
Report 188, Norwegian Institute of Technology, Trondheim.
Deere, D.U., 1981. Adverse geology and TBM tunnelling problems.
Proc. RETS, Society of Mining Engineers 1, 574586.
Deere, D.U., Deere, D.W., 1988. The RQD index in practice. In:
Proceedings of the Symposium of Rock Classication for Engineering Purposes, ASTM Special Technical Publication 984, pp.
9110.
Demuth, H., Beale, M., 1994. Neural Network Toolbox Users Guide.
The Mathworks Inc.
Duddeck, H., 1996. Challenge to tunnelling engineers. Tunnelling and
Underground Space Technology 11 (1), 510.
Einstein, H.H., Dudt, J.P., Halabe, V.B., Descoudres, F., 1992.
Decision aids in tunnelling principle and practical application.
Monograph, Swiss Fed. Oce of Transportation, Alptransit
Project.
Eisenstein, Z., 1999. Urban tunnelling challenges & progress. ITA 25th
Anniversary Commemorative Book.
Fausett, L., 1994. Fundamentals of Neural Networks. Architectures,
Algorithms and Applications. Prentice Hall International Editions,
New York.
Kavvadas, M., Hewison, L.R., Laskaratos, P.G., Seferoglou, C.,
Michalis, I., 1996. Experiences from the construction of the Athens

A.G. Benardos, D.C. Kaliampakos / Tunnelling and Underground Space Technology 19 (2004) 597605
Metro. In: International Symposium on Geotechnical Aspects of
Underground Construction in Soft Ground, pp. 277282.
Laughton, C., Nelson, P.P., 1996. The development of rock mass
parameters for use in the prediction of tunnel boring machines.
Eurock 96, 727733.
Lislerud, A., 1988. Hard rock tunnel boring: prognosis and cost.
Tunnelling Underground Space Technology 3 (1), 917.
McFeat-Smith, I., Tarkoy, P.J., 1979. Assessment of tunnel boring
performance. Tunnels and Tunnelling, 3337.
Menhrotra, K., Mohan, C.K., Ranka, S., 1997. Elements of Articial
Neural Networks. MIT Press, Cambridge, MA.
Nelson, P.P., 1993. TBM performance analysis with reference to rock
properties. In: Hudson, J. (Ed.), Comprehensive Rock Engineering,
vol. 4. Pergamon Press, New York, pp. 261291
(Chapter 10).
Okubo, S., Kfukui, K., Chen, W., 2003. Expert system for applicability
of tunnel boring machines in Japan. Rock Mechanics and Rock
Engineering 36 (4), 305322.
Peck, R., 1969. State of the art report: deep excavations and tunnelling
in soft ground. In: Proceedings of the Seventh International
Conference on Soil Mechanics and Foundation Engineering,
Mexico City, pp. 225290.

605

Sapigni, M., Berti, M., Bethaz, E., Busillo, A., Cardone, G., 2002.
TBM performance estimation using rock mass classications.
International Journal of Rock Mechanics and Mining Sciences
39, 771788.
Sharp, W., Ozdemir, L., 1991. Computer modelling for TBM
performance prediction and optimization. In: Proceedings, International Symposium on Mine Mechanization and Automation,
CSM/USBM, vol. 1, no. 4, pp. 5766.
Sietsma, J., Dow, J.F., 1991. Creating articial neural networks that
generalize. Neural Networks (4), 6779.
Sineld, J.V., Einstein, H.H., 1996. Evaluation of tunnelling technology using the decision aids in tunnelling. Tunnelling and Underground Space Technology 11 (4), 491504.
Tarkoy, P.J., 1981. Tunnel Boring machine performance as a function
of local geology. Bulletin of the Association of Engineering
Geology xvii (2), 4144.
Tarkoy, P.J., 1973. Predicting TBM penetration rates in selected rock
types. In: Proceedings of the Ninth Canadian Rock Mechanics
Symposium, Montreal.
Terzaghi, K., 1950. Geologic aspects of soft ground tunnelling. In:
Trask, J. (Ed.), Applied Sedimentation. John Wiley, New York
(Chapter 11).