Automated Data Acquisition and Monitoring

30

An Advanced Wireless System for Emergency Management in Hospitals

Angelo Biscotti1, Sara Spadoni1, Berardo Naticchia1 and Alessandro Carbonari2

1Polytechnic University of Marche, Smart Space Solutions SRL, Via Brecce Bianche, 60131 Ancona, Italy 2Polytechnic University of Marche, DACS Department, Engineering Faculty, Ancona, Italy. Tel: +390712204397; Fax: +39 0712204582; Mobile: +39 3402502768, alessandro.carbonari@univpm.it

Abstract

This paper reports preliminary results achieved during the experimentation of a new real time localization system, especially designed in support of hospitals’ emergency management. This system represents the first step of a wider ongoing research project, carried on by the authors in collaboration with the academic spin-off company Smart Space Solutions SRL, aimed at the developing of the W.E.M.A.S. (Wireless Emergency Management Aiding System) system, whose final aim is increasing hospitals’ efficiency in the management of the maxi emergencies consequent to disasters. WEMAS uses advanced wireless localization tracking and communication technologies to provide a number of features, such as: patient’s flows and patient’s healthcare progress tracking, equipment and medical personnel tracking, triage information management, ambulance scheduling and early communication of patient status from ambulance to the receiving emergency division. As a side effect, WEMAS system will equip hospitals with this new network infrastructure to optimize patient logistics for routine management. This paper describes also tests performed on the first release of WEMAS system, in the Emergency Division of the Hospital of Senigallia (Italy), demonstrating its features and technological feasibility.

Keywords: RTLS; wireless communication; disaster response; emergency management.

1. Introduction

This paper concerns the first step of a wider ongoing research project, named WEMAS (Wireless Emergency Management Aiding System), and carried out by the Department of Architecture Constructions and Structures (DACS) of the Polytechnic University of Marche, the academic spin-off company Smart Space Solutions SRL and the hospital of Senigallia.

The WEMAS, that will be developed to increase hospitals’ emergency divisions efficiency in the management of maxi-emergencies consequent to disasters, in its final release will provide the following functionalities: support to first aid management on disaster fields; support to patient transfer from disaster field to hospital; support to emergency division activities within hospital.

As a consequence, the WEMAS will have positive implications also in hospitals’ daily activities, allowing patients’ healthcare flow tracking, optimization of resource management and also facility tracking to prevent loss or robbery attempts.

The project is compliant to the Marche Region’s legislation that, by means of the regional act “DGR 49/2004”, imposes that every hospital must arrange a PEIMAF plan (“Piano di Emergenza Massiccio Afflusso Feriti”) devoted to coordination of hospital’s activities in case of massive victim afflux consequent to disasters.

Nevertheless, empirical experience has attested the inadequacy of the PEIMAF plan in maxi-emergencies management, because of the adoption of traditional communication and management instruments. Even the most recent research, carried out in countries very sensitive to catastrophic events, like Japan or USA, realized inadequacy of traditional means when coping with catastrophic events (Wickramasinghe et al. 2006). The last catastrophic event is represented by September the 11th attacks, when inefficient communication between disaster field and neighbor hospitals was experienced.

Nowadays, limits could be fortunately overcame by the advent of recent IC technologies, providing a valid support in hospital’s activities management (Chao et al. 2007), like faster patient’s identification, patients’ healthcare progress real-time tracking and so on. Some examples of such systems have been developed, tested and adopted in hospitals, but they all tend to be very invasive, moreover when installed in existing buildings.

This p

divisions, perf firstqualof S

The syfor fast an

2. State

PEIMAhospital fohospital’s

a)Figure 1

PEIMA

from theirFigure 1, victims’ catriage codcommunic

The indevelopmenvironmeRFID, UWsuch as:

1) hi2) in

For tha

3. The

Table relative topractices. patient coduring padocument

Insteadthrough a will be also

26th Internatio

paper presentscarried out byformance analt localization elity indices (th

Senigallia). ystem will be nd reliable com

e of the Art

AF standard por first aid; vdivisions.

1. Senigallia H

AF is applied r triage to furin case of disare process. Fde. Due to tcation betweenntroduction oent of innovents like hospWB) have bee

igh power connvasive installaat reason, the

New Approa

1 illustrates tho all the functi

For instance, ntaining his/htient’s healthcts and risks of d, according tolocal comput

o associated to

1.Holdin

4.Tempora

onal Symposium

s first studiesy the authors tlyses to discernexperimental t

hese tests have

based on Zigmmunication.

plans organizvictims transfe

Hospital Emergvictims

just in case ofrther medical saster Senigal

Figure 1-b alsothe involvemen personnel, n

of Radio Freqvative real-tipitals. These en installed an

nsumption, reations of big sWEMAS pro

ach Brought

he results of aionalities that check-in wou

her ID, triage care progress

f document loso the WEMAter unit or a Po a bracelet w

ng room

2.Triage 3.Treati

ary hospitalizat

m on Automati

s, aimed at dthrough the fon the system’s tests, to perfoe been carried

gBee technolo

e rescues opeer form disast

gency Divisios paths in case

f massive patiecare (Figure 1

llia hospital emo shows that ent of all honot achievablequency Identime managemsystems, base

nd employed i

quiring connesized devices aoject wants to

about by Ou

analyses carriecould be reas

uld change it code and hea. The main liss.

AS, patient’s IDPDA, passing worn by the pa

ing rooms

tion

ion and Roboti

31

eveloping a following stepsmain function

orm the reliabout in one of

ogy, as it prov

erations on a ter scene to h

b)on organizatione of disaster m

ent flux and r1). As it can bmergency divvictim’s flows

ospital divisioe through stanification technment systemsed upon varion few hospita

ection to the eand cabling. release a new

ur W.E.M.A.S

ed out togethesonably provistandard proc

alth status), in imit of the s

D and data wilthrough emer

atient and con

ics in Construc

first WEMASs: nalities; bility of the syf the most dif

vides such a ti

three level sthospital; victim

n in daily routmanagement (b

equires to fullbe noticed fro

vision occupies follow diffeons, it requirndard communnologies couls, applied atous communials. However t

electric system

product.

S.

er with the hoided by the Wcedures (assignfavor of a batandard pract

ll be recordedrgency divisiotaining a Zigb

ction (ISARC 2

S system relea

ystem, throughfficult environm

iny structure,

tructure: orgam managemen

tine (a) and reb).

ly re-organize om comparinges new areas, rent courses ies a constannication protold give a gret different cication protocthey all presen

m;

ospital of SenWEMAS for au

nment of a padge to be autotice lays in th

d in the centran’s intra-netw

bee End-Devic

009)

ase for emerg

h the use of sments: the ho

considered cr

nization of a nt and care w

-organization

rescue operatg the two parinvolving thein relation to

nt and continocols. at impulse toomplex opercols (WiFi, IRnt the same li

nigallia’s persoutomating hoaper form to eomatically updhe frailty of p

al hospital datawork. The samce dispositive,

gency

signal spital

ritical

field within

of

tions, rts of em in

their nuous

o the rative RDA, imits,

onnel, spital every dated paper

abase me ID

, able

to communicdevices will apatients and h

Query- - - - - -

Alert - - - - -

In addition

progress, by expertise scou

Figure 2. Tra

1.Figure

4. The WE

The WEM802.15.4 stanfor use in emuseful proper

ZigBee's cseveral advanworks in ISMis easy to dep

cate with portalso be providhealthcare reso

Taby

Patient’s idenHealthcare plClinical paraReal-time locaTimeling Available res

Patient’s suppPatient’s cliniPatient’s drifRobbery attemHealthcare p

n, by means oconnecting t

uting and cons

a) aditional triage

23. The proced

EMAS Techn

MAS providesdard medium

mbedded applicrties to be impcurrent focus ntages provide

M band; it is reloy; it allows v

Automate

table and fixeded to doctorources’ positio

ble 1. The WE

ntification lan

ameters cation

source

port call ical parameter mft mpts rogress tracking

of their PDAso patients’ brsulting, as sho

e paper (a); Pr

2.dure for “Med

nology

s its features m access and Z

cations requirplemented in h

is to define ed by the Zig

eliable and selfvery long batt

ed Data Acquis

3

ed wireless ders, nurses andons.

EMAS’ featur

monitoring

, doctors will racelets. Thro

own in Figure

b) rogrammation

3.dical expertise

through hardZigbee stack cring low data rhealth care woa general-purgbee protocof healing; it sutery life; it is a

sition and Mon

32

evices of a ned hospital ass

es foreseen foCheck out

- Ad- Dim- Self

Health car- Tr

Request - Me- Me- Busy- Cal

be able to visough their de3.

n and Assignat

4scouting and

dware and socommunicatiorates and low

orkplaces (Lorrpose, cheap, l, that make

upports large nsafe and low

itoring

etwork installets, allowing

or the final reletdmission mission lf-dimission re riage code assigna

edical expertise sedical test or instsy ll monitoring

sualize patientevices they co

tion of patient

4.consulting” th

oftware compon protocol. T

power consurincz et al. 200self-organizinit strategic fo

number of nocost technolo

led in the hofor real time

ease.

ation

scouting and construment reservati

t’s identity andould also ask

t’s identificatio

5.hrough medic

onents basedThese protocoumption, and p06). ng mesh networ hospital enodes (16 bit adogy.

spital. Portabmonitoring o

sulting ion

d his healthcarfor a medic

on bracelet (b)

cal PDAs

on the IEEols are intendepresent a lot o

work. There arnvironments: ddress space);

le of

re cal

).

E ed of

re it it

26th International Symposium on Automation and Robotics in Construction (ISARC 2009)

33

The WEMAS exploits three Zigbee device types (Figure 4-a), compliant to standard application profile: each one presenting specific features, as depicted in Table 2.

Our WEMAS devices are equipped with the integrated radio and microcontroller Chipcon/Texas Instruments CC2431 chip with 128K flash memory. CC2431 has an embedded localization hardware engine, which estimates the position of a node by evaluating the received signal strength (RSSI) attenuation (Michel et al. 2006), according to the Path Loss Model (PLM) formula:

( )( )dLognARSSI 1010 ⋅⋅+−= (eq. 1) where d represents the distance in meters from the sender node, while A is the received signal in dBm at

a distance of one meter with a sending power of 0dbm. The decentralized localization algorithm is performed entirely by the “blind” node, reducing the

communication costs and bottlenecks. A mobile node who wants to evaluate its position sends a broadcast message to all ZRs of the network in its radio range; each ZR replies with a short message whose payload contains its fixed coordinates and the message’s RSSI received, measured in dBm (see Figure 4-b). At least three RSSI indications are then written in the location hardware registers with ZRs coordinates, A and n values of the PLM formula in eq. 1. Output contains the coordinates of the evaluated position. The procedure to evaluate A and n values is described in (Chipcon/Texas Instruments 2007).

Table 2. ZigBee device types with their features

Zigbee Coordinator (ZC) Zigbee Router (ZR) Zigbee End-Device (ZED)

• one and only one required for each ZB network;

• initiates network formation;

• acts as 802.15.4 PAN coordinator (FFD);

• not necessarily dedicated device, can perform applications.

• may associate with ZC or with previously associated ZR;

• acts as 802.15.4 coordinator (FFD);

• local address (destination) allocation/de-allocation;

• participates in multi-hop routing of messages.

• mobile node;

• shall not allow association;

• shall not participate in routing;

• put to sleep by parents.

The engine is not capable to evaluate node’s position if less than three ZRs reply to the broadcast

request. In this case the position can be evaluate through gate control. The utilization of a dedicated hardware reduces considerably the delay of the mathematical process. The Zigbee standard profile suggests to have main powered ZR, in order to maintain the radio transceiver in listen mode and have the possibility of routing messages throughout the network. The necessity of using cables dramatically decreases the flexibility of the network in existent hospital environments, because fixed nodes have to be placed in proximity of an electrical source. However the Zigbee specifications allow to use a synchronous communication protocol which uses a special type of message called beacon.

The beacon is a sync message, which defines a time interval for message passing, but increases extremely the latency, in order to minimize duty cycle. WEMAS system uses a new approach, developed by the academic spin-off Smart Space Solutions srl. This proprietary protocol is intended for reduce ZR’s duty cycle in asynchronous mode. It allows the arrangement of full wireless networks, because battery powered, requiring minimum maintenance efforts, with a 2 years long power life.

Each ZR node remains in sleep mode for reducing power consumption until it has to route a message or reply to a request. In these cases an asynchronous signal wakes it up from sleep and allows it to work without waiting for a sync.

Figure 4The localiz

healing proptechnologies.

RFID hasnetworks. Zigtechnologies. spectrum) tecemploying a techniques detaken by the Ieveryday use i

Infrared scommunicatiomultipath effemessages canconnection is

UWB allowWi-Fi boa

strength of aobstructions s

Some envi

Figure 5. S

5. Prelimin

As previonecessary to dis measured th

Preliminaris to say the ea), whose thic

a) . example of azation hardwaerty, make t

s a lower ranggbee compliaIn (ZigBee A

chnology, thetechnique, ge

escribed aboveIEEE in the 8in real environsystems needon capabilitiesfects, thanks tn be sent into too low, the mws very accurasts a higher a Zigbee devsuch as doors ironments nee

Signal attenuat

nary Experim

ously explainedevelop a reliahrough RSSI rry tests were pemergency divckness ranges

Automate

a mesh Zigbeeare engine andthe WEMAS

ge, compared want devices inAlliance 2007 802.15.4 phyenerally knowe, a ZigBee de802.15.4 standnments with r to have a ls, so it can bo application several direcmessage is lik

rate positioninrange, but th

vice is propoor walls (Figu

ed to increase

tion in corresp

ments on the

ed, in order table wireless sreceived by evperformed in vision of the Sbetween 0.4 a

ed Data Acquis

3

e network (a); d the low pow

extremely su

with costs of nstead aren’t s7) it is showeysical layer pr

wn as frequenevice may find

dard is known real data trafficline of sight to

be used just foof mesh netw

ctions. If one ely to direct it

ng, but it is verhe infrastructuortional to thure 5). the number o

pondence of d20

Wemas Rea

to assure thasensor networvery ZR withinthe worst env

Senigallia hospand 1.2 m.

sition and Mon

34

RSSI based mwer communicuitable for h

readers, and subject to siged how in adrotocol increancy division md itself sharingas carrier sensc bear prove Zo communicaor gate controworking. Sincepath is lost f

t along anothery expensive aure is expensihe transmissio

of ZR to main

different kind007).

al-time Local

at all the WErk (i.e. reliablen radio range)

vironment whepital, built with

itoring

b)

message exchacation capabilihospitals with

interference pgnificant interddition to DSases the oppomultiple accesg a channel wise multiple acZigBee’s robuate one anothol. Instead Zie multiple patfor some reaser path. and complex. ive and needs

on power and

ntain redundan

s of obstructio

lization Syste

EMAS’ featuree communicati). ere WEMAS h a heavy mas

ange for localizities, combine

h respect to

problems withrference with SSS (direct seortunities for ss (FDMA). ith interfererscess (CSMA).

ustness. her and haveigbee devices ths exist in a sons or the lin

s electric powd decreases i

ncy and avoid

ons (OneRF T

em

es will work ion among all

could be likelsonry structure

zation (b). d with the selother trackin

h existing LANother wireles

equence spreacoexistence bEven with th. The approac Both tests an

e not compledo not suffermesh networnk budget of

wer. The signin presence o

bottlenecks.

Technology,

properly, it l the nodes tha

ly installed, the (see Figure 6

lf-ng

N ss ad by he ch nd

ex rs k, a

nal of

is at

at 6-

a)

Figure 6.

Senigal

made wallorganizatio

FollowEmergencnoticed a sixteen ZR

nine six Z one

The phbe noticedthe walls w

aFigure 7. W

InsteadFigure 8-asystem. Lo

It has received sthan -95 d

26th Internatio

)

Ground level

llia Hospital Els (Figure 6-b)on makes loca

wing first prelicy Division. FZC device ha

R devices have ZR in the twZR in six diffe ZR (no. 20) nhotographic red that this sugwith reversible

a)

WEMAS insta

d ZED devica, in order to ocalization havto be outline

signal always ldBm than it is

onal Symposium

l plan of the Sbric

Emergency D), and partitionalization a touminary experi

Figure 7 illustas been placedve been deploywo corridors (nerent rooms alnext to the doeport in Figur

ggested technoe means (like g

allation layout

es have been verify that all ve been perfoed that our slimited betwee neglected; ot

m on Automati

Senigallia Hospck wall (b) and

Division is a thning walls bui

ugher task. iments, the fintrates fixed dd close to the yed as followsno. 4, 5, 6, 7, 1long the corridoor of one roore 7-b depicts ology requires glue or similar

t in Senigallia Hof some dev

moved acrosthe room inc

ormed using thsensor deviceen -95 dBm atherwise if it is

ion and Roboti

35

pital Emergend of a partition

hree level buililt with hollow

nal testing havevices positiomiddle of the

s: 0, 12, 13, 21, 2

dors (no. 1, 14m. some of the rrather small s

r systems).

b)Hospital Emevice positioninss nine differ

cluded within the algorithm ds hold a builand -40 dBm s higher than

ics in Construc

b)

ncy division (aning wall (c).

lding, charactw-solid bricks

ve been carrieoning throughe left side corr

22) ; , 15, 16 17, 23

routers installsized devices,

) ergency Divising (b). ent positions the emergency

detailed in parlt-in mechanis(that is to sa-40 dBm, tha

ction (ISARC 2

c)

a); details of a

terized by ver(Figure 6-c).

ed out at the ghout the strucridor, next to

3);

ed on the hosproperly pack

ion (a), and ph

during testiny division canagraph 4. sm for RSSI

ay, if the recein it is outputt

009)

)

0.4 m thich s

ry thick solid Its cramped s

ground level octure: as it cathe left wall, w

spital’s walls: ikaged and fixe

hotographic re

ng, as illustraten be tracked by

evaluation oived RSSI is lted at -40 dBm

solid

brick space

of the an be while

it can ed on

eport

ed in y our

f the lower m). A

good commuwhen the ZED

Experimenexamples of tstrongest ZR the blind nod

As guide-lsubdivided byarea. ZRs mu

a)

Figure 8. Va

6. Conclus

Performanthat our WEMaffluence, or optimized resthrough the in

First localhaving very h

Within littTable 1, will experimental

References

[1] C. Chaotechnolo

[2] ChipconLocation

[3] K. Lorintracking"

unication qualiD is located inntal tests gavthe performedwas identified

de (i.e. ZED) wline for furthey heavy walls,

ust also be inst

arious location

sions

nce analyses rMAS can offein hospital’s

source and constallation of lization exper

heavy structuretle time, a probe installed monitoring.

, W. Chen, Y.ogy", Int J Elen / Texas Instrn Engine" <Uncz, M. Welsh ", Pres Ubiqui

Automate

ity is assured n every positio

ve back good d localization d as the closewas obtained. er installations, can be monitalled next to a

ns occupied b

relative to maer in order todaily activitie

ommunicationthe full-wireleriments demoes. ototypal versioand tested in

Chy, B. Lin (ctron Healthcruments (2007

URL:http://fo(2006), "Mot

it comput., Sp

ed Data Acquis

3

if RSSI valueon of the builresults fromtests: in every

st and a good s it was also nitored by usinall door openi

Po

Rea

Exb)

by the ZED deestimated po

axi-emergencieo become a stes. In the last n managementess, low invasivonstrated the

on of WEMASn the emergen

(20079, "Imprc, 3(2): 175-927), "System-oncus.ti.com/lit

teTrack: a robupringer-Verlag

sition and Mon

36

e higher than ding.

m this point oy case the rec

d match betwe

noticed that: ang 16 ZRs, things, in order

sition n.1

al position

xtimated position

evices during tositions (b, c).

es managementrategic instrucase, the ado

t, besides impve system, bareliability of

S system, endncy Departme

roving patient2 n-Chip for 2.4/ds/symlink/ust, decentrali

g London Lim

itoring

-85 dBm is m

of view. Figurceived RSSI ween the real an

a 350 m² larghat is to say oto avoid ZED

Po

R

n

Ec)

testing (a); com.

nt allowed idement either inoption of the proving structsed on our unthe WEMAS

dowed with soent of Senigal

t safety with R

4 GHz ZigBee/cc2431.pdf>ized approach

mited, DOI 10

measured by at

res 8-b and 8was higher thand the estimat

e built area, sne ZR per 20

D localization

osition n. 4

Real position

Extimated positio

mparison betw

entifying a sern the case of WEMAS can

ture’s safety lenique technoloS, even in cas

ome further fullia Hospital

RFID and MO

e®/ IEEE 80

h to RF-based 0.1007/s00779

t least one ZR

8-c show soman -85dBm, thted position fo

surrounded an0 m² large buierrors.

on

ween real and

ries of featuref massif victimn be useful foevel, achievabogy. se of building

unctionalities ofor continuou

OBILE

02.15.4 with

location 9-006-0095-2.

R,

me he or

nd ilt

es ms or le

gs

of us

26th International Symposium on Automation and Robotics in Construction (ISARC 2009)

37

[4] D. Michel, D. Toggenburger (2006), "Sensor network soccer", Diploma Thesis WS, University of Applied Sciences Rapperswil HSR, Department of Electrical Engineering, Institute for Communication Systems.

[5] OneRF Technology (2007), “B2400MC-uTiny Technical Manual” <URL:http://backoffice.inware.it/files/prodotti/download/onerf/manuale%20zigbee%201.pdf>

[6] N. Wickramasinghe, R.K. Bali, R.G. Naquib (2006), "Application of knowledge management and intelligent continuum for medical emergencies and disaster scenarios", Conf. Proc. IEEE “Eng Med Biol Soc.”, 1:5149-52

[7] Zigbee Alliance White Paper (2007), "ZigBee and Wireless Frequency Coexistence", < www.zigbee.org/imwp/download.asp? ContentID=11745>