DATA TRANSFER LINK 2: 3D VISUALISATION CLIENT AT CUSTOMER PREMISES

H. Guijt(1)

, K. Debeule(2)

(1) TERMA, Schuttersveld 9, 2316XG Leiden, The Netherlands, tel: +31 71 524 0835, fax: +31 71 514 3277, email:

(2)

ESA/ESTEC, Keplerlaan 1, 2201AZ Noordwijk, The Netherlands, tel: +31 71 565 5345, fax: +31 71 565 3911,

email:

1 ABSTRACT

During environmental tests of a spacecraft in a vacuum

chamber a large amount of measurement data is

continuously generated and made available for real-time

monitoring and analysis, or further elaboration. In the

frame of DTL1 (Data Transfer Link 1) a data

communication capability between ESTEC Test Centre

and customers’ premises was created such that test data

were made available in near real-time to test engineers

located remotely.

The objective of DTL2 (Data Transfer Link 2) is to

enhance the visibility of test data by means of Virtual

Reality techniques. For instance, search and position

identification of a specific sensor on board of the test

specimen (usually out of several hundreds of sensors)

can be quickly done. Evolution of the measured

physical quantity (e.g. temperature) can be visualized or

processed for further analysis.

At customers’ premises an interface website allows

downloading of several files to enable the installation of

a 3D presentation client for the test actually in progress.

As far as the ESTEC Test Centre is concerned, all the

measurement data collected in STAMP (System for

Thermal Analysis, Measurement and Power supply

control), including test specimen position and chamber

properties, can be rapidly transmitted and visualized at

different remote locations.

Starting from the system performance requirements

specified for DTL 2, this paper will present the main

features of DTL 2 and the current status of operability.

Figure 1: the virtual reality presentation

2 THE NEED FOR REMOTE MONITORING

The goal of remote monitoring is to reduce the cost and

increase the flexibility of thermal tests. It achieves these

goals by allowing thermal experts to work from the

comfort of their own office, i.e. by removing the need

for them to travel to the Test Centre to witness the test

in person. Instead the test data is brought directly to

their desktop, where it can be observed in real time, and

processed using any of the usual in-house tools, even

while the test is still running.

Moreover, since there are no restrictions on the remote

monitoring client, it is possible to observe the test from

any location that has an internet connection (and for

places that don’t have an internet connection, the remote

client can work in offline mode with any data it has

received previously).

3 SECURITY ASPECTS

In order to be able to work, remote monitoring requires

a connection between the data handling system and the

outside world. This raises two security concerns:

1. Since the data handling system plays a critical

role in the thermal testing process, it is vital

that it cannot be compromised by an outside

attack.

2. Data transmitted by the remote monitoring

system may not be intercepted or altered by

unauthorized 3rd

parties.

DTL2 employs a range of techniques to address these

concerns. Figure 2 shows the main components of the

security system:

Figure 2: security setup

The “master server” (on the left) is responsible for

conducting the test within the Test Centre. To protect it

from outside intrusions, firewall 1 is configured to only

allow connections to be created from inside the Test

Centre. Thus it is possible for the master server to

initiate communication with the teletest server, but

connections in the opposite direction are always rejected

by the firewall. In this way, even if the teletest server is

compromised somehow, the master server is protected

from outside interference.

The “teletest server” (in the centre) is responsible for

the remote monitoring system. It contains a copy of the

test configuration and data, allowing these to be

displayed on the remote clients. The teletest server itself

is protected by a second firewall, “firewall 2”. This

firewall is configured to only allow incoming

connections for the remote monitoring system.

Moreover, such incoming connections are required to

present a trusted certificate (which are created by the

Test Centre on a per-customer basis). If the remote

client cannot present a trusted certificate, or if the

presented certificate is not a valid certificate that was

signed by the Test Centre, the connection attempt is

rejected.

If the trusted certificate of an incoming connection is

found to be valid, the remote client must then present a

username and password. These determine which subset

of the data on the teletest server the remote client is

allowed to see; access to thermal tests, spacecraft

models, and specific sensors is all determined on a per-

user basis.

Finally, all traffic between the teletest server and the

remote client is encrypted using strong encryption,

making it impossible to intercept or alter the data that is

being exchanged.

4 DATA VISUALISATION USING VIRTUAL

REALITY TECHNIQUES

The remote monitoring client uses virtual reality

techniques to increase the visibility and overview of the

test. It does this by displaying a 3D model of the

spacecraft and test facility, and projecting known

information (such as sensor values, spinbox rotation

angles, solar state, etc.) onto the model. The user can

then navigate around the model in real time.

Figure 3 shows a spacecraft mounted inside the Large

Space Simulator at ESTEC. The coloured dots indicate

the locations of sensors; the scale on the right correlates

the colours of the dots to the actual temperature of each

sensor.

Although it is possible to obtain exact numeric values

for each sensor simply by hovering the mouse over it,

the strength of the virtual reality presentation lies in the

overview it offers over the total state of the test:

Figure 3: virtual reality presentation

� The orientation of the spacecraft within the test

facility is immediately visible.

� The overall thermal state can be observed at a

glance.

� The “find mode” highlights sensors with a

specific name or number (shown in the

previous image as an arrow).

� New operators will find it easy to get up to

speed on the positioning of the sensors by

examining the 3D model.

� The 3D model makes it easy to understand

which part of the spacecraft is visible to the

solar beam and which is not.

� The 3D model is a useful entry point to more

detailed presentations (such as graphs or

tabular displays). Any node of the spacecraft

model may be selected for visualisation in

another style of presentation; for example, to

display all sensors on a specific instrument as a

graph, just select the instrument in the 3D

display and choose “show as graph” from its

context menu.

4.1 Model reuse

The virtual reality presentation relies on having a 3D

model of the spacecraft (3D models of the facility are

available at ESTEC). Since creating 3D models is

expensive, DTL2 makes it possible to reuse existing

mechanical or thermal models of the spacecraft. These

models are converted using an off-the-shelf tool that can

convert models in a variety of formats (including

CATIA, STEP, Pro*E, and many others).

One potential problem, especially with mechanical

models, is that they may be extremely detailed and

therefore contain a very large number of polygons.

While modern VR workstations have no problem

displaying extremely complex models (one of the

models we tried has ~20 million polygons and displayed

fine), older machines may have trouble handling very

large models. In order to maintain an acceptable

rendering speed, the remote monitoring client has the

ability to reduce the number of displayed polygons in

real-time. Thus visualisation on slower machines is also

supported, with minimal loss of rendering fidelity.

4.2 Sensor placement

In addition to the model itself, the locations of the

sensors must also be entered into the remote monitoring

client. Sensor locations are entered using a simple point

and click mechanism: to place a sensor, select it from

the list of sensors and click in the 3D window on the

location of the model where the sensor is located. Using

this method, sensors can be accurately placed on any

part of the model. Sensors need to be placed only once;

after the sensor definitions have been entered into the

configuration database, all users of the test can use

them. To avoid accidental movement of sensors, sensors

can only be placed by specifically authorized users.

Typically this authorisation is revoked after the sensors

have been placed, and before the test starts.

After the sensors have been placed, it is possible to

create 2D line drawings of the model from any

viewpoint. These line drawings clearly indicate the

locations of the sensors on the spacecraft (and on the

facility, if so desired). See Figure 4 for an example:

Figure 4: 2D drawing showing sensor locations

The figure shows the Large Space Simulator (with one

part cut out to allow a view of the spacecraft), and all

sensors that are visible from this viewpoint.

4.3 Visualisation modes

The main visualisation mode of the virtual reality

presentation is to display temperatures as coloured dots.

However, it is also possible to display the equilibrium

state of each sensor. The equilibrium state is defined as

the absolute value of the difference between the average

value of the sensors over a recent period, compared to

the average value of the sensor over a period in the past,

and provides a measure of the thermal stability of the

system.

Moreover, in either mode it is possible to display the

sensors on a neutral grey background, or to colour each

part by the average of the values of the sensors located

on that part. Figure 5 shows how this works for the

equilibrium values:

� Grey parts do not have any sensors located on

them, and thus remain grey.

� Green, strongly visible parts have a high

equilibrium value (i.e. a high degree of thermal

instability).

� Green, almost transparent parts have a low

equilibrium value (i.e. are thermally stable).

Thus the virtual reality presentation shows at a glance

which parts are thermally stable and which are still

fluctuating.

Figure 5: equilibrium display

It is also possible to combine both modes, using

transparency to indicate the equilibrium value and

colour to indicate the temperature.

It is not necessary to display the entire model all the

time: parts of the model may be selectively made

transparent or invisible with the click of a button.

Multiple, different models may be kept in memory,

although only one is displayed at any given time.

With the correct hardware, the virtual environment can

be displayed as a stereo image, giving a much greater

sense of depth than what is available on a 2D display.

5 CONNECTION WITH OTHER

PRESENTATION STYLES

Apart from the new virtual reality presentation, the

remote monitoring client also offers all the usual

presentations available within STAMP. Among others,

these include:

� Graphs: STAMP has powerful, highly versatile

graph presentations that can display any

number of curves over any length of time

(memory permitting), with multiple vertical

scales, choice of linear / logarithmic modes,

powerful scaling options, markers, plotted

against time or against other sensors.

� Tabular: these presentations show the values of

many sensors, either over time or at a single

timestamp.

� Alarm: generates alarms if sensors exceed their

warning limits, alarm limits, or delta limits, or

if their predicted value exceeds their alarm

limits.

� Connector: transmits data on a socket to

another application. This offers a simple way

to connect the remote presentation client to

other tools at remote sites.

� Excel: creates .CSV files ready for loading into

Excel (or other tools).

� Equilibrium: determines the equilibrium values

of a group of sensors.

� Prediction comparison graph: allows predicted

values from the thermal analysis software to be

compared in real time to actual measured

values obtained from the test.

� Waterfall graph: shows the evolution of a

group of sensors over time.

All STAMP presentations support both real-time and

archived data mode where possible.

Figure 6: other presentations in the remote client

6 USER FRIENDLINESS

Although user-friendliness was already a design goal for

STAMP, for the remote presentation client it is

especially important since it is intended to be used by

remote users who do not have easy access to Test

Centre operators for asking questions.

Several measures were taken to accommodate these

users:

� The user interface of the remote presentation

client was overhauled to a significant degree to

produce the maximum possible degree of

clarity. This includes such changes as

removing jargon, adding unobtrusive help

messages, and changing screen layouts to be

identical between different presentation styles.

� An easy to use installer was created that

installs the remote presentation client with just

a few mouse clicks.

� Online, context-sensitive help was added to all

presentations.

Together these measures go a long way towards making

the remote presentation client accessible to novice users.

7 APPLICABILITY TO OTHER (ESTEC)

TEST FACILITIES

The only thing that is facility-specific about the virtual

reality presentation or the remote monitoring client are

the models of the test facility. However, like the

spacecraft models, these can be loaded into the system

on an as-needed basis. Thus, to support different test

facilities (such as the new Phoenix chamber at ESTEC),

all that is needed is a test facility model.

8 INITIAL EXPERIENCES

The virtual reality presentation was used for the first

time during the Herschel test campaign at ESTEC in

January / February 2007. Although it was still in

prototype form at this stage the users were enthusiastic

about the possibilities of the new system.

The remote monitoring feature was used for the first

time during the SMOS test campaign at ESTEC in April

2007 (this was also the second use for the virtual reality

presentation). Data was presented remotely in Madrid

during the course of the test, receiving a similar

enthusiastic response.

No problems were encountered during either test,

although a few very good suggestions for further

evolution of the system were received.

In both cases, we were able to reuse existing models (an

engineering model for Herschel and both an engineering

model and a thermal model for SMOS). Placing the

sensors was performed onsite at ESTEC, and took about

one hour for Herschel (placing 60 sensors – note that

most of this time was spent locating them in the

Herschel documentation!) and about two and a half

hours for SMOS (placing 460 sensors).

9 CONCLUSIONS

9.1 Remote monitoring

Remote monitoring is an effective way to decrease the

cost and increase the flexibility of thermal testing. It

allows the customer to bring in extra experts on

demand, without incurring the cost of flying them in or

keeping them onsite at the Test Centre all the time.

Similarly, it offers a middle road between having

operators onsite at all times and unsupervised testing:

operators can check the progression of the test remotely,

only coming in if a problem is spotted during test

execution.

9.2 Virtual reality presentation

Virtual reality is a great way to offer a good overview of

the total state of the spacecraft and chamber. It is not a

replacement for graphs and tables (nor is it intended as

such), but it augments those presentation styles with

extra information that can be understood very quickly.

Moreover, since it offers those other presentation styles

through a “drilldown” interface, it acts as an effective

stepping stone towards obtaining a detailed

understanding of the current state of the test.

10 ACKNOWLEDGEMENT

The authors of this paper would like to thank Astrium

(Herschel) and CASA (SMOS) for allowing the use of

their spacecraft models, and for investing the time

needed to actually use the virtual reality presentation.

Furthermore, we would like to thank J. van der Meulen

(ETS) for his invaluable help for setting up the

machines and software for both test campaigns.

The virtual reality presentation is based on an earlier

development called “VRAIV”, which was developed in

the frame of an ESA contract. VRAIV is a virtual reality

simulation tool geared towards simulating AIV

processes. It is described in one of the other papers

submitted for this conference.

Terma is currently in the process of commercializing

STAMP for other customers. For more information,

please contact H. Guijt ( ).