Page 33 - profile-ok
P. 33

研究群   |   Research Laboratories









 Network System and Service


 Laboratory







 Research Faculty  Group Profile


 Jan-Ming	Ho  The	Network	Systems	and	Services	Laboratory	has	seven	principle	investigators:	Ling-Jyh	Chen,	  Intelligent  Monitoring,  Alert  and  Notification Another	 major	 empha-  short.	For	this	reason,	one	wants	to	build	them	from	independently
 Research	Fellow  Meng-Chang	Chen,	Sheng-Wei	Chen,	Tyng-Ruey	Chuang,	Jan-Ming	Ho,	Jane	Liu,	and	Der-Nian	  sis	of	our	research	involves	methods	and	tools	for	ensuring	that	the	  developed	components	and	have	them	run	in	an	open	environment
 Electrical	Engineering	and	Computer	Science	,	  Yang.	Our	team	collaborates	on	several	joint	projects	under	the	common	theme	of	user-centric	  symbiotic	system	consisting	of	a	user-centric	device	and	its	user(s)	  on	commodity	platforms.	RARMS	aims	to	make	this	way	of	building
 Northwestern	University  cyber-physical	services.	We	also	address	several	technical	issues,	including	mobile	computing,	  works	safely	(i.e.,	it	never	does	any	harm	and	all	unavoidable	errors	  time-sensitive	robotic	applications	possible.
 Ling-Jyh	Chen  social	computing	and	quality	of	experience.	Our	research	results	have	been	published	in	major	  are	either	recoverable	or	tolerable).	This	problem	is	particularly	criti-    ●  PLASH: Platform for Location Aware Service with Human Com-
 Assistant	Research	Fellow  conferences	and	journals.	We	also	participate	in	the	development	and	transfer	of	technologies	to	  cal	since	most	SISARL	devices	are	semi-automatic,	and	may	rely	on	  putation
 Computer	Science	,	University	of	California	at	Los	  industry.	A	summary	of	our	current	activities	is	provided	below.  the	user(s)	to	perform	some	mission-critical	functions.	This	problem
 Angeles	       is	made	more	challenging	by	the	fact	that	typical	users	are	untrained,	  The	PLASH	project	started	in	mid	2009.	It	takes	a	novel	approach	of
 1. User-Centric Cyber-Physical Services
 MengChang	Chen  and	their	skills	vary	widely	among	the	user	population	and	for	an	in-  exploiting	 human	 computation	 for	 wireless-networked,	 location-
                dividual	user	over	time.
 Research	Fellow  In	a	cyber-physical	system	(CPS),	computing	devices,	sensors,	and	actuators	coordinate	with	  aware	applications.	Human	computation	uses	massive	groups	of	net-
 Computer	Science	,	University	of	California,	Los	  each	other	and	interact	closely	with	the	physical	environment.	CPS	is	a	new	area	of	research	  We	are	exploring	the	idea	of	having	the	device	monitor	user	actions	  worked	human	beings	and	intelligent	things	to	solve	problems	that
 Angles  that	fuses	together	embedded	and	real-time	systems,	sensor	devices	and	networks	and	hybrid	  at	runtime.	When	the	device	detects	user	actions	that	may	lead	to	  computers	cannot.	A	location	aware/based	application	makes	use	of
 control	systems.	User-centric	(UC)	devices	and	systems	are	used	at	the	user’s	discretion,	do	  information	on	the	geographical	position(s)	of	mobile	device(s)	to
 Sheng-Wei	Chen  what	the	user	wants	them	to	do,	and	work	according	to	the	user’s	expectation.	Examples	in-  violations	of	the	underlying	assumptions	for	its	correct	operations,	it	  provide	services	that	are	impossible	without	location	awareness.	Ex-
 Assistant	Research	Fellow  clude	many	automation	and	assistive	devices,	and	wireless-networked	location-aware	devices	  works	with	the	user	to	correct	the	situation	and	prevent	misuses	that	  amples	include	the	delivery	of	information	on	the	whereabouts	of	a
 Department	of	Electrical	Engineering	,	National	  may	have	serious	consequences.	Thus	far,	we	are	attacking	this	prob-
 Taiwan	University  and	services	for	improving	quality	of	life	and	enhancing	societal	wellbeing.	On-going	projects	  lem	in	an	ad	hoc	manner,	one	device	at	a	time.	We	need	to	be	able	to	  friend,	or	alerts	about	new	bumps	or	accidents	on	the	road.	We	want
 in	this	area	include	UCAADS	(User-Centric	Automation	and	Assistive	Devices	and	Systems)	and	  translate	general	principles	into	design	guidelines	and	rules,	as	well	  to	explore	how	human	computation	can	enhance	the	quantity	and
 Tyng-Ruey	Chuang  PLASH	(Platform	for	Location	Aware	Services	with	Human	Computation).	  quality	of	information	for	such	services,	and	how	to	build	services
 Associate	Research	Fellow  as	working	methods	and	tools.            that	are	impossible	without	human	computation.
 Computer	Science,	New	York	University    ●  User-Centric Automation and Assistive Devices and Systems  Technology Transfer Efforts Our	work	on	UCAADS	has	prepared	us	for
 Jane	W.	S.	Liu  The	goal	of	UCAADS	research	is	the	advancement	of	technologies	for	building	low-cost,	high-  two	advanced	development	efforts:	MeMDAS	(Medication	Manage-  We	are	currently	constructing	a	multi-layer	middleware	to	support
                                                                     smart	mobile	device	platforms,	including	Andriod	and	iPhone.	In	par-
 Chair	Research	Fellow  quality	personal	and	home	automation	and	assistive	devices	and	systems.	We	call	these	devices	  ment,	Dispensing	and	Administration	System)	and	RARMS	(Robotic	  ticular,	our	platform	aims	to	facilitate	the	use	of	human	computation
 Electrical	Engineering,	Massachusetts	Institute	of	  and	systems	collectively	SISARL	(Sensor	Information	Systems	for	Active	Retirees	and	Assisted	  Application	Resource	Management).	The	RARMS	project	is	funded	by
 Technology  Living),	which	is	the	name	of	the	Academia	Sinica	thematic	project	in	which	this	work	began.	  ITRI.	MeMDAS	is	a	collaborative	effort	with	the	National	Taiwan	Uni-  by	 diverse	 location	 aware/based	 application	 systems	 and	 services,
                                                                     and	 thus	 simplify	 the	 complexity	 of	 such	 application	 systems.	The
 De-Nian	yang  Examples	of	SISARL	include	smart	medication	dispensers,	autonomous	home	appliances	and	  versity	Hospital	(NTUH)	and	ITRI	Mechanical	and	Systems	Research	  Communication	Layer	enables	the	use	of	various	wireless	technolo-
 Assistant	Research	Fellow  robotic	helpers	designed	to	help	elderly	individuals	stay	well	and	independent.	Other	SISARL	  Laboratories.	 It	 has	 been	 partially	 supported	 by	 NTUH	 since	 2008.	  gies	(e.g.,	WLAN,	WiMax,	and	3G)	and	communication	structures	(e.g,
 Department	of	Electrical	Engineering	,	National	  are	automation	and	point-of-care	tools	for	enhancing	the	quality	and	reducing	the	costs	of	  Since	July	2009,	it	is	also	partially	funded	by	the	one-year	經濟部科  Vehicle	to	Infrastructure,	Vehicle	to	Vehicle,	or	Vehicle	to	Mobile	de-
 Taiwan	University  health	and	medical	care.	Below	is	an	overview	of	our	current	research	and	advanced	develop-  技研究發展專案.	  vice).	The	Data	Layer	is	for	data	representation,	storage,	and	access.
 ment	efforts.	Details	on	these	efforts	and	links	to	open	source	software	can	be	found	at	the	  MeMDAS	is	a	system	of	smart	medication	cabinets,	mobile	nursing	  The	Service	Layer	aims	to	provide	the	basic	building	blocks	for	serv-
 project	homepage	(http://SISARL.org).  carts,	and	software	tools	for	the	purpose	of	preventing	medication	  ices/applications.	This	layered	architecture	allows	application	build-
 Middleware and Tools	A	common	requirement	of	SISARL	is	that	such	a	device	must	be	flexible,	  dispensing	and	administration	errors.	One	initial	result	is	an	intelli-  ers	to	conveniently	create	their	systems	by	simply	using	PLASH	API’s
 i.e.,	easily	configurable,	customizable,	and	capable	of	adapting	to	changes	in	the	user’s	needs	  gent	nursing	cart	(iNuC).	In	addition	to	tools	for	bar-code	controlled	  to	access	the	Service	Layer.
 and	skills.	We	have	adopted	the	workflow	paradigm	to	achieve	greater	flexibility	and	used	  medication	 administration	 and	 automatic	 patient	 record	 keeping,	  As	case	studies	and	for	proof-of-concept	and	experimentation	pur-
 workflows	in	two	innovative	ways.	One	way	is	to	use	workflows	as	behavior	specifications	of	  iNuC	provides	its	users	with	modern	work	and	time	management,	in-  poses,	we	have	built	several	location-aware	applications	that	use	hu-
 devices,	user	actions,	and	device-user	interactions.	In	essence,	these	specifications	are	execut-  telligent	monitoring	and	alert	tools,	and	a	web	portal	to	the	hospital	  man	computation.	One	example	is	“map’n	track	friends”	(http://plash.
 able	device	and	user	models.	We	use	them	to	simulate	devices	and	device-user	interactions	for	  information	system.	iNuC	version	1.0	is	now	ready	for	experimental	  iis.sinica.edu.tw/Danny/*.action),	which	can	run	on	both	Android	and
 assessing	the	usability	of	devices	and	ensuring	that	the	symbiotic	system	of	device	and	user	  use	by	NTUH	nursing	staff.	We	are	currently	implementing	software	  iPhone	platforms.	As	its	name	implies,	Map’n	track	friends	is	a	loca-
 works	as	desired.	SISARL	Simulation	Environment	was	designed	for	this	purpose.	  components	that	integrate	iNuC	with	other	components	of	MeMDAS,	  tion	service	that	lets	you	keep	track	of	the	whereabouts	of	friends,
                including	a	multi-user	medication	station	server	that	manages	medi-
 We	also	adopt	the	workflow-based	design	and	architecture	for	embedded	devices	and	robots,	  cations	in	cabinets	under	its	control	and	enforces	bar-code	control-  family,	and	colleagues	on	a	map.	The	system	localizes	you	only	when
 using	a	workflow	engine	as	middleware	for	the	dynamic	integration	of	reusable	components.	  led	medication	dispensing.	The	capability	of	supporting	concurrent	  the	application	is	running.	The	system	is	secure	via	a	two-way	hand-
 The	light-weight	Embedded	Workflow	Framework	(EMWF)	was	built	and	is	continuously	en-  dispensing	by	multiple	users	from	the	same	station	will	be	a	distin-  shake	validation	system:	You	must	invite/request	friends	to	be	added
 hanced	for	this	purpose.	We	aim	to	make	the	enhanced	version	of	EMWF	suitable	for	a	wide	  guishing	feature	of	MeMDAS.	The	other	distinguishing	characteristic	  to	your	list	and	they	must	then	accept	your	requests.	You	will	be	able
 range	of	SISARL	devices	and	systems.	We	are	incorporating	mechanisms	for	end-to-end	re-  is	flexibility.	The	system	is	designed	to	be	easily	configured	to	sup-  to	see	only	friends	who	have	accepted	your	requests,	and	vice-versa.
 source	and	quality	of	service	management,	which	will	make	EMWF	capable	of	supporting	real-  port	 centralized,	 distributed,	 and	 hybrid	 dispensing	 processes	 and	  Another	example	is	an	application	that	allows	voluntary	users	to	up-
 time	system	abstractions	needed	by	time-critical	cyber-physical	systems	and	services.  easily	customized	by	individual	departments	and	hospitals	to	follow	  load	pictures	taken	while	traveling.	The	system	can	use	the	collected
                their	own	protocols.	The	MeMDAS	software	system	and	tools	are	built	  pictures	and	users’	traveling	experiences	as	basis	to	recommend	tour
                around	a	workflow-based	architecture.	Their	implementation	on	the	  routes	to	other	drivers	based	on	their	locations	and	previous	travel
                Microsoft	.NET	Workflow	Framework	(WF)	provides	us	with	an	invalu-  routes.		Finally,	the	cyber-physical	social	network	system	VProbe	de-
                able	case	study	on	the	use	of	workflow-based	designs	for	complex	  scribed	later	in	the	section	on	social	sensing	applications	is	an	ex-
                SISARL.	We	plan	to	replace	WF	with	our	own	enhanced	version	of	  ample	of	novel	applications	that	exploit	the	synergy	between	human
                EMWF.	                                               computation	and	location	awareness.
                                                                  2. Mobile Computing Research
                The	RARMS	project	started	in	March	2010.	Its	goal	is	to	establish	a	firm
                foundation	 for	 a	 framework	 of	 real-time	 middleware	 and	 resource	    ●   Networked Sensing Systems
                management	services.	RARMS	specifically	targets	service	and	social
                robots	(i.e.	household	automation	devices,	delivery	robots	and	tour	  Our	research	focus	is	on	networked	sensing	systems	in	hostile	envi-
                guides,	physical	therapy	assistants,	and	so	on).	For	these	applications,	  ronments	in	which	network	contacts	are	intermittent,	an	end-to-end
                it	is	important	to	keep	development	costs	low	and	time-to-market	  path	between	the	source	and	the	destination	rarely	exists,	disconnec-
 32                                                                                                              33
   28   29   30   31   32   33   34   35   36   37   38