BLINDTRACK: Guiding System for Visually Impaired

Locating System for Running on a Track

Ferdinand Kemeth

, Sven Hafenecker

Agnes Jakab

e Varga

, Tam

as Csielka

and Sylvie Couronn

Fraunhofer Institute for Integrated Circuits, Nuremberg, Germany

Ateknea Solutions, Budapest, Hungary

Keywords:

Real-time Locating System, Angle of Arrival, Round-trip Time, Visually Impaired Runners, Guiding System.

Abstract:

Visually impaired people need to renounce several social activities what the sighted people can enjoy. In this

paper we refer to the project BLINDTRACK which has the major goal to develop a guidance system. A

real-time locating system (RTLS) based on radio signals guides the runners with the highest level of safety

by estimating the angle of arrival (AoA) and round-trip time (RTT). First results show the position accuracy

of the proposed locating system with real-world data. BLINDTRACK provides an enormous freedom for

the visually impaired runners in compare to the other solutions: Sighted and blind runners will have the

opportunity to perform sport together without another person’s assistance.

1 INTRODUCTION

In recent years the EU and all its member countries

have committed themselves to create a barrier-free

Europe. There are still many obstacles preventing

people with disabilities from fully exercising their

fundamental rights including their Union citizenship

rights and restricting their participation in society on

an equal basis compared with non-disabled people.

Those rights include the right to free movement, to

choose where and how to live, and to have full access

to cultural, recreational, and sports activities. Regard-

ing sports, particular running activities, visually im-

paired do not have equal possibilities.

The most common way of running is the so-called

guided running in that blind runners train and race

with a sighted runner with the help of a tether (Amer-

ican Foundation for the Blind, 2014). With the

BLINDTRACK project the Consortium aims to raise

the level of accessibility of visually impaired to sport

to reﬂect the need for an effective assistive technol-

ogy which would facilitate the well-being of visually

impaired while decreasing their exclusion from sport

and leisure activities. The proposed system helps vi-

sually impaired people integrate to the community

with increasing conﬁdence, better health condition

and higher tolerance level of sighted people.

The aim of BLINDTRACK project is to develop

a running facility embedded to a 400 m athletic track

for visually impaired people to run without another

person’s assistance. BLINDTRACK will be able to

bring signiﬁcant changes in training opportunities for

visually impaired. The objective is to increase the

number of blind and partially sighted athletes with

the creation of a tailored infrastructural facility that

can be the ﬁrst step to train without sighted volun-

teers. Although BLINDTRACK focuses on visually

impaired users, the system provides online available

training results to the sighted people: This enhances

the market opportunities and further development for

small and medium enterprises (SME).

In this paper we focus on the project’s locating

part as follows: Section 2 provides an overview for

the proposed system and the project organization.

Section 3 gives basics on the used locating techniques

and Section 4 shows the locating performance of the

system under development. The last section summa-

rizes the paper together with the next steps.

(European Commission, 2010) (United Nations,

2007) (England Athletics Limited, 2012) (United

States Association of Blind Athletes, 2014) (Com-

petitor.com, 2014)

183

Kemeth F., Hafenecker S., Jakab Á., Varga M., Csielka T. and Couronné S..

BLINDTRACK: Guiding System for Visually Impaired - Locating System for Running on a Track.

DOI: 10.5220/0005143001830189

In Proceedings of the 2nd International Congress on Sports Sciences Research and Technology Support (icSPORTS-2014), pages 183-189

ISBN: 978-989-758-057-4

 2014 SCITEPRESS (Science and Technology Publications, Lda.)

2 BLINDTRACK OVERVIEW

The BLINDTRACK system structure is intended to

built up three main components:

• the BlackFIR (Fraunhofer IIS, 2012) unit which is

a high precision locating system,

• the belt unit which is a vibro tactile belt and

• a Central Control Unit (CCU).

2.1 System Description

The BlackFIR has the responsibility to locate the ath-

letes on the track in real time, while the belt unit co-

ordinates the athletes with vibrations while running

and helps them to avoid the obstacles and the colli-

sion situations. The CCU predicts and calculates the

different trajectories for each athlete and also ﬁlters

and singles out the right commands to be sent.

Figure 1 shows the general architecture of the

complete BLINDTRACK system. It is necessary to

create an own local BLINDTRACK network on the

athletic ﬁeld, which is completely separated from the

local area network. This separation ensures to keep

the BLINDTRACK server safe, prevents the unde-

sired system overloads and shutdowns and makes the

operation fast and continuous. The server has a Linux

operating system and a runtime environment to fulﬁl

the controlling tasks that includes the calculation of

the trajectories, the selection and transfer of the com-

mands. The CCU has also a client side, where one or

more computers can connect to the BLINDTRACK

server as client device. The client computer has a

graphical user interface (GUI), which helps the oper-

ator who is the supervisor of the athletic ﬁeld and the

runners. The GUI is also responsible for the adminis-

trative issues and the statistical data display. All these

tasks are not required real time conditions therefore

they are detached from the BLINDTRACK server.

The communication protocol between the mod-

ules is based on novel wireless technologies where

the high speed and reliability are crucial parameters.

The communication with the belt unit will be imple-

mented by Wi-Fi access points that provide easy and

fast information transmission, and prevents the inter-

ference with the BlackFIR locating system. The belt

consists of a driver module and an actuator part. The

driver module is responsible for the communication

towards the CCU and for the control of the belt ac-

tuators. This module also contains a power supply

unit, which drives the complete belt and a compass

module which determinates the relative orientation of

each actuator to the north. The ﬁrmware of the belt

has the responsibility for the translation of the rela-

tive to north commands to the actuator. This can min-

imize the intensity of wireless communication in the

BLINDTRACK system.

BlackFIR is an innovative radio frequency (RF)

based locating system, which is able to operate in

real time and detects the exact position of each run-

ner on the track. The BlackFIR system consists of

several antenna units, mobile transponders carried by

athletes and a central unit. The communication be-

tween the antenna units and the central unit occurs via

Ethernet. The mobile transponder comprises a special

transceiver chip to locate and identify the athletes.

Figure 1: General system architecture of the BLIND-

TRACK.

2.2 Project Organization

BLINDTRACK project management structure was

designed and agreed by the consortium members to

ensure an effective and straightforward project man-

agement mechanism. The main principles were to

set up and maintain an organisational structure that

ensures the highest level technical and ﬁnancial im-

plementation of the project and the efﬁcient exploita-

tion of the project results and effective decision mak-

ing structure. The partners represent several Euro-

pean countries such as Germany (PPS and Fraun-

hofer IIS), Hungary (Ateknea Solutions, BSK, IN-

FOALAP), the Netherlands (Elitac), Norway (Adap-

tor), Spain (Eneso, IBV) and their professional expe-

riences cover all the neccessary ﬁelds that are needed

to conclude in a successful project result by the end

of 2015. A more detailed partner description is found

in the Appendix.

3 ADAPTED LOCATING SYSTEM

The BlackFIR system developed at Fraunhofer IIS

consists of four mounted receiver units, upto 30 mo-

bile units (transponders) and a central computing unit.

icSPORTS2014-InternationalCongressonSportSciencesResearchandTechnologySupport

184

The communication between the receiver units and

the central unit occurs via Ethernet. For identiﬁca-

tion and locating the nanoLOC transceiver chip (Nan-

otron Technologies GmbH, 2007) is used. The in-

frastructure measures the angle of arrival (AoA) and

the round-trip time (RTT) using RF signals generated

by the mobile units in the 2.4 GHz ISM frequency

band. Using these two locating techniques every re-

ceiver unit estimates the direction and distance of ac-

tive transponders and sends the results to a central

unit, where the position is determined by combining

all available locating information.

3.1 Basics of AoA

Adaptive antenna arrays allow estimating the angle of

arrival of a received signal (Tuncer et al., 2009). The

received signal source has to be in a far ﬁeld condi-

tion. Far ﬁeld condition results in signal propagation,

which is nearly ﬂat. Figure 2 shows the phase alter-

ation on different channels of an antenna array.

...

ant.1

Δ ψ

ant.2 ant.3 ant.n

Figure 2: Received far ﬁeld signal at the antenna array.

To estimate the angle of arrival, different high res-

olution algorithms like Multiple Signal Classiﬁcation

(MUSIC), estimation of signal parameters via rota-

tional invariance techniques (ESPRIT) or the Min-

imum Variance Cappon-Beamformer (Tuncer et al.,

2009) can be applied. These algorithms use the co-

variance of the different received signals to estimate

the angle of arrival. BlackFIR uses a MUSIC al-

gorithm with an additional forward backward spatial

smoothing (FBSS) which improves robustness against

multipath non-line-of-sight (NLOS) signals. The re-

sult of the MUSIC algorithm is an angle spectra like

shown in Figure 3; the maxima of the spectra show the

estimated angle of arrival. The resulting peaks of the

MUSIC spectra get narrower with an additional pre-

processed FBSS. Inherent with this, an improvement

of multipath-robustness against NLOS-signals is also

realized.

3.2 Basics of RTT

The measurement of distances between a tag (”X”)

and given receiver positions (”1”, ”2” and ”3”) allows

0 20 40 60 80 100 120 140 160 180

−2

Angle of Arrival [degree]

P MUSIC

P MUSIC fbss

Figure 3: MUSIC (power) with and without forward back-

ward spatial smoothing.

the calculation of the tag position, as shown in Fig-

ure 4. BlackFIR uses the nanoLOC transceiver chip

to implement round-trip time (RTT) (Std 802.15.4a,

2007) based distances. The distance is measured by

a so called two way ranging procedure. In this pro-

cedure time of ﬂight (TOF) is measured by anchor to

tag response time and vice versa. This gives the op-

portunity to compensate differences in the frequency

reference of the tag and the anchor.

2x t2

2x t1

2x t3

Figure 4: RTT-based calculation of position.

Another well-known method to measure distances

is to calculate time of arrival (TOA) or time differ-

ence of arrival (TDOA). For TOA it is necessary to

have synchronisation between receiver and transmit-

ter which is often not possible in an adequate way.

Due to that TDOA is a possibility, where only the in-

frastructure is synchronized such as in satellite based

global positioning systems (GPS). Using TDOA does

not require an extra reply signal which lowers the

channel usage and minimizes the measurement time.

However RTT requires no synchronisation techniques

between the receiver units, what inﬂuences the system

infrastructures as well as increasing the cost.

3.3 Infrastructure Setup

The BlackFIR locating system consists of a receiver

unit estimating the AoA and the RTT. In order to cal-

BLINDTRACK:GuidingSystemforVisuallyImpaired-LocatingSystemforRunningonaTrack

185

culate a two-dimensional position, it is necessary to

use at least one of these receiver units. In practice

the data is affected by multipath effects, fading and

noise based error. To improve the estimates, more

units are used at one scenario in a time duplex sequen-

tial measurement procedure as shown in Figure 5. All

available sensor data are collected and combined by a

central server were the position is then calculated by

either a Kalman ﬁlter (Kalman, 1960) or particle ﬁlter

(Sanjeev et al., 2002) based algorithms.

angle 3

AOA / RTT Receiver 1

AOA

RTT

Receiver 2

AOA

RTT

Receiver 3

TAG

2 ε_angle

2 ε_distance

distance 2

Figure 5: Principal abstract of BlackFIR RF locating.

4 EVALUATION

4.1 Measurement Campaign

In order to validate the performance values, a mea-

surement campaign was executed in the N

urnberg sta-

dium. The used test ﬁeld is a standardised (DIN

18035-1, 2003) type ”A” all-weather running track

with eight lanes and a length of 400 m as shown in

Figure 6.

Figure 6: All-weather running track used for testing.

Aim of the measurement campaign:

• Identify the performance of the system in the tar-

get environment.

• Evaluate the optimal infrastructure structure.

• Optimize the positioning algorithms.

To analyse the current system performance in dif-

ferent circumstances, the following scenarios were

explored:

• Walking Scenario. A person was walking along

the third running track. The sender was afﬁxed

at his inner shoulder. The scenario was repeated

three times.

• Running Scenario. A person was running along

the third running track. Likewise the walking sce-

nario, the sender was ﬁxed at his inner shoulder.

• Scenario with a Bicycle Rider. A person was

riding a bicycle along the third running track. The

sender was carried in the outer hand.

To get reference information for the positions an

optical highly accurate system called iGPS (Nikon

Metrology, 2010) was used. The optical component

was mounted on a helmet worn by the athlete as

shown in Figure 7. IGPS ensures point locating ac-

curacy down to 200 µm.

Figure 7: iGPS probe mounting at the top of a helmet.

During the different scenarios various datasets of

sensor information were recorded to analyse and im-

prove the locating system.

The types of raw information are:

• In-phase and quadrature components of the RF

signal received by the antenna unit

• Measured distances between mobile tag and re-

ceiver

• Estimated angle of arrival at the receiver units

• Battery life-time information

Four receiver units were placed arround the run-

ning track (one unit for every 45-degree bend) in or-

der to ensure line-of-sight to the transponders and for

coveraging the whole test ﬁeld.

icSPORTS2014-InternationalCongressonSportSciencesResearchandTechnologySupport

186

4.2 Analysis

In Figure 8 and Figure 9 the measured raw angle of

arrival and raw distance data over the time are shown.

Each signal of the different receivers is plotted in an-

other colour.

5.442 5.444 5.446 5.448 5.45 5.452 5.454 5.456

x 10

100

150

200

Time [sec]

Angle [degrees]

Receiver 1

Receiver 2

Receiver 3

Receiver 4

Figure 8: Angle results based on Angle of Arrival estima-

tion.

5.442 5.444 5.446 5.448 5.45 5.452 5.454 5.456

x 10

100

150

200

Time [sec]

Distance [m]

Receiver 1

Receiver 2

Receiver 3

Receiver 4

Figure 9: Distance results based on round-trip time.

Every single curve shows spikes depending on

channel propagation, fading and multipath effects

(e.g. Receiver 1 in Figure 8 at 5.451 · 10

sec). Be-

cause of the dynamic scenario caused by the runner’s

movement, these effects can be detected and compen-

sated later.

Straight lines are representing gabs in the infor-

mation caused by a poor wireless connection or inter-

ferers (e.g. Receiver 4 in Figure 8 in the time interval

between 5.448 and 5.451 · 10

sec). Due to the sys-

tem’s high update rate and redundancy, it is not nec-

essary to get information from every receiver unit at

all times.

An exemplary view of the actual measured dis-

tances and estimated angles is presented in Figure 10:

Each of the four receiver units is showing the esti-

mated angle as a straight line towards the sensors’ po-

sition. The circles around them are representing the

measured distances. In best cases, all lines and circles

are crossing in one single point: the athlete’s position.

The iGPS reference system was installed in the

right part of the stadium due to the need of reference

values. The estimated system latency as well as the

accuracy of the position is determined by the data ob-

served in that part of the stadium. In contrast to that,

the data rate was determined over all data stored in the

corresponding scenario.

Figure 11 shows the position trajectory of the

Runner‘s Position

Receiver Unit

Estimated Angle

Estimated Distance

Figure 10: Position measurement with angle and distance

estimation.

−100 −50 0 50 100

−50

x−Axis [m]

y−Axis [m]

Reference

Calculated

Figure 11: Measured position in the scenario with a running

person.

measured position values (green line) in the scenario

with a running person for two laps. The red line repre-

sents the position given by the iGPS reference system

which could only cover the right part of the running

track. The position estimates have a small variance

from the reference position. In addition, the estimates

at the right part of the stadium have better perfor-

mance compared with position estimates from the left

part. Reason: The calibration process was performed

in the right part, where the iGPS system was installed,

so the left antennas were not calibrated as ﬁne as the

right antennas. The measurement results of the two

laps are both similiar to each other and comparable to

the described effects.

The estimated accuracy of the position is deter-

mined by the data that where observed in that part

of the stadium, where the iGPS reference system was

installed due to the need of reference values. In con-

trast to that, the data rate was determined over all data

stored in the scenario. Table 1 summarizes the mea-

sured system performance:

Table 1: Summary of measurement results.

Estimated latency 440 msec

Data rate 6.6 Hz

Mean deviation (MAE) 0.689 m

Maximum deviation 2.60 m

”Estimated latency” describes the delay between

the athlete’s actual and calculated position caused by

wireless transmission, network trafﬁc and proccess-

BLINDTRACK:GuidingSystemforVisuallyImpaired-LocatingSystemforRunningonaTrack

187

ing time for the algorithms. ”Data rate” describes the

update rate of the estimated positions; the update rate

for the beld commands sent to the athlete will be less.

The distribution of the measured position errors is

shown in Figure 12.

Figure 12: Distribution of the position error in meters.

5 SUMMARY & ROADMAP

In the scope of this paper, we presented an adaptation

for the BlackFIR locating system to the project needs

of BLINDTRACK. For a system using RTT and AoA

technology, we showed detailed results on a measure-

ment campaign and the accuracy of the calculated po-

sitions. Our system gives visually impaired people

the opportunity to enjoy several social activities with

simple installation complexity. Compared to visual

tracking system, we are able to identify and track ev-

ery runner, even in extreme situations, where the run-

ner is surrounded by many others. In next steps, ad-

justments will be applied resulting in increasing the

accuracy of the current system and further cost reduc-

tion. At the end of this project, the complete system

will be tested and evaluated with blind and non-blind

runners.

ACKNOWLEDGEMENTS

The project has received funding from the European

Union Seventh Framework Programme (FP7/2007-

2013) under the grant agreement No. 605821. It is

a 24 months research project that was launched on 1st

of January 2014 with the cooperation of nine partners.

We thank Mohammad Alawieh, Fraunhofer IIS,

for supporting and revising the paper.

REFERENCES

American Foundation for the Blind (2014). Tips for

runners with visual impairments. http://www.afb.org/

info/living-with-vision-loss/recreation/running-2805/

235.

Competitor.com (2014). Visually impaired runners ready

to tackle boston. http://running.competitor.com/2014/

04/news/visually-impaired-runners-ready-tackle-

boston 98512.

DIN 18035-1 (2003). Sports grounds - part 1: Outdoor play

and athletics areas, planning and dimensions.

England Athletics Limited (2012). Running & sprint-

ing with guides. http://metroblindsport.org/

downloadlibrary/guide-running-v2.pdf.

European Commission (2010). European disability strat-

egy 2010-2020: A renewed commitment to a barrier-

free europe. http://eur-lex.europa.eu/LexUriServ/

LexUriServ.do?uri=COM:2010:0636:FIN:EN:PDF.

Fraunhofer IIS (2012). BLACKFIR 2.4 brochure.

Kalman, R. (1960). A New Approach to Linear Filtering

and Prediction Problems. Journal of Basic Engineer-

ing, 82 (series d) edition.

Nanotron Technologies GmbH (2007). nanoNET Chirp

Based Wireless Networks White Paper.

Nikon Metrology (2010). iGPS datasheet: Priciples of op-

eration / Scalebar / i5 Probe kit / Dynamic tracking

kit.

Sanjeev, Arulampalam, M., Maskell, S., Neil, G., and

Clapp, T. (2002). A Tutorial on Particle Filters for

Online Nonlinear/Non-Gaussian Bayesian Tracking.

IEEE Transactions on Signal Processing, vol.50 edi-

tion.

Std 802.15.4a (2007). IEEE Standard for Information Tech-

nology - Telecommunications and Information Ex-

change Between Systems - LANs and MANs - Speciﬁc

Requirements - Part 15.4: Wireless MAC and PHY

Speciﬁcations for LR-WPANs - Amendment 1: Add Al-

ternate PHYs. IEEE Std 802.15.4aTM-2007 edition.

Tuncer, T., Friedlander, B., and Yasar, T. (2009). Classical

and Modern Direction-of-Arrival Estimation. Aca-

demic Press.

United Nations (2007). Convention on the rights of per-

sons with disabilities. http://www.un.org/disabilities/

default.asp?navid=13&pid=150.

United States Association of Blind Athletes (2014).

Sports adaptations. http://usaba.org/index.php/sports/

sports-adaptations/.

APPENDIX

SME Participants

INFOALAP-Informatics for the Visually

Impaired Foundation - Hungary

INFOALAP is actively involved in system speciﬁca-

tions, based on its experience with IT for visually im-

paired, and supports the technology optimization by

providing direct evaluation feedback for. They also

support hardware and software development by the

evaluation of ergonomic aspects by its low vision IT

engineers, as well as being active in system tests.

icSPORTS2014-InternationalCongressonSportSciencesResearchandTechnologySupport

188

PPS GmbH - Germany

PPS is the supplier of the technology that is capable of

real-time detection of the running people in the run-

ning track. It closely cooperates with Fraunhofer In-

stitute IIS during the research period and supplies the

prototype of the project. PPS deﬁnes the prototype in

regard of later products and assists the development

partners in deﬁning the real use cases as well as the

general evaluation of the developments.

ELITAC - Netherlands

Besides supervising, ELITAC’s role comes at the

phase of tactile belt research and manufacturing.

They strongly cooperate with IBV while they are

working on the development of the tactile belt. ELI-

TAC has already developed tactile devices but none of

them was for the blind.

Eneso Tecnolog

ıa de Adaptaci

on S.L. - Spain

ENESO is a distributor of BLINDTRACK system.

They have a deep knowledge of the accessibility mar-

ket in Spain, so they will be very active in promoting

and placing the system. They will also contribute their

experience by testing and validating the product.

ADAPTOR HJELPEMIDLER AS - Norway

ADAPTOR plays a signiﬁcant role in the consortium

as the employer of visually impaired and distributor of

assistive products and provides the research partners

with tangible information on the needs of the market

and assist ﬁeld testing.

RTD Participants

ATEKNEA Solutions Hungary Kft. - Hungary

Ateknea Solutions brings together four research and

innovation companies operating at a European level

for over 15 years. The group pools the expertise and

know-how of more than 130 professionals working

in ﬁve different locations: Barcelona, Brussels, Bu-

dapest, Krakow and Valetta. The innovative com-

panies have successfully participated in more than

150 different projects ﬁnanced by the European Com-

mission. ATEKNEA is the coordinator of the project

and responsible to ensure ﬂuent project ﬂow.

Fraunhofer Institute for Integrated Circuits IIS -

Germany

The Institute has a ﬁrst prototype research result,

called RedFIR® know-how and experience that is a

state-of-the-art wireless tracking technology that lo-

cates people and objects in real time and with high

precision. Its main role is to select the best ﬁtting lo-

calization technology for a successful project. Com-

pared to current video-based approaches, this radio-

based technology offers a major advantage: its track-

ing capability is not diminished by obstacles obscur-

ing the line of sight. The RedFIR® real-time tracking

system is more responsive, accurate and ﬂexible than

any comparable technology. Position data is made

available in fractions of a second and automatically

analysed using pattern recognition and event detec-

tion methods. User-speciﬁc data preparation and vi-

sualization modes are provided in real time. It has an

accuracy of a few centimetres, making event detection

results and automatically generated statistics highly

reliable. In the consortium Fraunhofer will develop

a system based on RedFIR® but fulﬁlling the special

needs of visually impaired people and ﬁnd out a so-

lution that is still affordable and marketable with not

forgetting the basic need: maximum accuracy with

minimum price.

IBV - Biomechanics Institute of Valencia - Spain

IBV do research to understand the tactile sensing

and the perception mechanism in different conditions.

Measurements are taken to deﬁne the sensitivity of the

skin from the density and the intensity point of view

as well. They aim to ﬁnd the optimal sensing posi-

tions on the perimeter of the trunk to feel directions

with conﬁdence and deﬁne the number of the vibrat-

ing elements on the belt, and their control to guide a

person to the direction we intend to. They have to co-

operate closely with Elitac in order to manufacture a

deﬁned number of prototypes for testing. A continu-

ous cooperation with the control and communication

circuit designer RTD during the control circuit devel-

opment is also essential.

Other

Budapesti Sportszolg

altat

o K

ozpont - Hungary

During the project preparation and implementation

Budapesti Sportszolg

altat

o K

ozpont help with prac-

tical advices of blind running behaviour and critical

points of their secure training. In the testing phase

it will make the ﬁeld available and will actively take

part in the validation and dissemination.

BLINDTRACK:GuidingSystemforVisuallyImpaired-LocatingSystemforRunningonaTrack

189