TABLE 3.

Results of the localization using our method and AMCL. The predicate succesfull is yes when the robot localizes accurately with respect its ML hypothesis error stays localized for the remainder of the path. The distance needed to acquire localization is both expressed as actual localization of ML and as self-reported localization (by assessing convergence). Note that self-reported localization can be reported erroneously (i.e.,. false confidence). The average ML error after localization is also reported and the average number of hypotheses used both before and after self-reported localization has been achieved.

	Scenario	A		B										C
	Descr	Init Region		Total map										Clutter
	Path	A1	A3	B1	B2	B3	B4	B5	B6	B7	B8	B9	B10	C11	C12
ours	succesfull	yes	yes	yes	yes	yes	yes	yes	yes	no	yes	yes	yes	yes	no
	actual dist ML	1.7	4.9	1.7	8.3	4.9	5.3	1.1	7.2	n/a	5.5	7.1	10.3	16.4	n/a
	self-rep dist	13.2	4.9	16.3	22.0	8.4	24.1	15.7	20.9	16.9	5.5	4.0	6.0	2.9	9.1
	ML avg er aft con	0.05	0.33	0.07	0.23	0.28	0.36	0.13	0.25	n/a	0.24	0.51	0.16	0.08	n/a
	max # hyp bef con	200	200	200	200	200	200	200	200	200	200	200	200	200	200
	max # hyp aft con	7	9	7	10	7	5	33	11	n/a	8	7	5	6	n/a
AMCL	succesfull	5/5	5/5	2/5	0/5	5/5	0/5	0/5	0/5	3/5	4/5	4/5	3/5	2/5	5/5
	actual dist ML	5.2	6.2	3.9	n/a	9.3	n/a	n/a	n/a	6.5	10.3	1.2	6.3	13.1	8.0
	self-rep dist	5	6.9	6.4	n/a	13.3	n/a	n/a	n/a	8	12.2	2.9	7.4	12.9	8.7
	ML avg er aft con	0.19	0.32	0.18	n/a	0.29	n/a	n/a	n/a	0.2	0.17	0.25	0.36	0.42	0.22
	max # hyp bef con	5e4	5e4	5e4	n/a	5e4	n/a	n/a	n/a	5e4	5e4	5e4	5e4	5e4	5e4
	max # hyp aft con	5e4	44217	5e4	n/a	5e4	n/a	n/a	n/a	17378	28235	13527	5e4	5e4	5e4