WP4-16 - Revision history

Acorde: /* Improvements */

2023-03-13T15:44:16Z

Improvements

← Older revision		Revision as of 15:44, 13 March 2023
Line 56:		Line 56:

	Another aspect addressed in the context of WP4 is an '''analysis to assess if state-of-the-art deep learning methodologies might provide feasible and advantageous solutions''' for embedded targets. Specifically, the possibility to enhance via AI/ML a currently heuristic functionality integrated '''on the multi-baseline attitude estimation algorithm of GLAD+''' was assessed.		Another aspect addressed in the context of WP4 is an '''analysis to assess if state-of-the-art deep learning methodologies might provide feasible and advantageous solutions''' for embedded targets. Specifically, the possibility to enhance via AI/ML a currently heuristic functionality integrated '''on the multi-baseline attitude estimation algorithm of GLAD+''' was assessed.
	The result of these assesment was very promising, since with a limited effort (produce a synthetic dataset, as explained later, design the NN and translation to an embedded target), it was possible to get ~~'''~~interesting results in accuracy (without any domain expertise), and an embedded implementation able to work at real-time''.		The result of these assesment was very promising, since with a limited effort (produce a synthetic dataset, as explained later, design the NN and translation to an embedded target), it was possible to get interesting results in accuracy (without any domain expertise), and an embedded implementation able to work at real-time.

Acorde: /* Improvements */

2023-03-13T15:43:40Z

Improvements

← Older revision		Revision as of 15:43, 13 March 2023
Line 55:		Line 55:
	Beyond these graphically apparent example, a systematic analisys on a set of flights of different nature lead to 15%/23.3% average improvements for horizontal/vertical accuracy in static conditions, and 11.9%/31.2 in dynamic conditions.		Beyond these graphically apparent example, a systematic analisys on a set of flights of different nature lead to 15%/23.3% average improvements for horizontal/vertical accuracy in static conditions, and 11.9%/31.2 in dynamic conditions.

	Another aspect addressed in the context of WP4 is an '''analysis to assess if state-of-the-art deep learning methodologies might provide feasible and advantageous solutions''' for embedded targets. Specifically, the possibility to enhance via AI/ML a currently heuristic functionality integrated '''on the multi-baseline attitude estimation algorithm of GLAD+''' was assessed (~~more~~ details in D4.6 of C4D).		Another aspect addressed in the context of WP4 is an '''analysis to assess if state-of-the-art deep learning methodologies might provide feasible and advantageous solutions''' for embedded targets. Specifically, the possibility to enhance via AI/ML a currently heuristic functionality integrated '''on the multi-baseline attitude estimation algorithm of GLAD+''' was assessed.
			The result of these assesment was very promising, since with a limited effort (produce a synthetic dataset, as explained later, design the NN and translation to an embedded target), it was possible to get '''interesting results in accuracy (without any domain expertise), and an embedded implementation able to work at real-time''.


			Further explanation on the evaluation method are given below, while details are explaines in D4.6 of C4D.

	Last, but not least, in the framework of WP4, ACORDE has also provided a "Mavlink interface" to ease SW, and therefore GLAD+ system integration (more details in D4.8).		Last, but not least, in the framework of WP4, ACORDE has also provided a "Mavlink interface" to ease SW, and therefore GLAD+ system integration (more details in D4.8).

Acorde: /* Description */

2023-03-13T15:38:40Z

Description

← Older revision		Revision as of 15:38, 13 March 2023
Line 36:		Line 36:

	[[File:wp4-16_06.png\|frame\|center\|Basic architecture of the multi-gnss fusion sensor functionality of GLAD+]]		[[File:wp4-16_06.png\|frame\|center\|Basic architecture of the multi-gnss fusion sensor functionality of GLAD+]]


	==Improvements==		==Improvements==

Acorde: /* Validation */

2023-03-13T15:38:16Z

Validation

← Older revision		Revision as of 15:38, 13 March 2023
Line 69:		Line 69:
	After the integration test, ACORDE, with the support of CATEC, set up ''test flight'' for different fligth condtions, considering both, maneuvres that could challenge the positioning system, and feasible operations according CATEC experience (free flight). Those flights took place 2023/01/12.		After the integration test, ACORDE, with the support of CATEC, set up ''test flight'' for different fligth condtions, considering both, maneuvres that could challenge the positioning system, and feasible operations according CATEC experience (free flight). Those flights took place 2023/01/12.

	<gallery widths=500px heights=500px caption="Some instants of the ''test ~~flight~~'' ">>		<gallery widths=500px heights=500px caption="Some instants of the ''test flights'' ">>

	File:Wp4-16-03.png\|center\| Drone used for the tests		File:Wp4-16-03.png\|center\| Drone used for the tests

Acorde: /* Validation */

2023-03-13T15:37:19Z

Validation

← Older revision		Revision as of 15:37, 13 March 2023
Line 81:		Line 81:

	The following figure shows the set up designed for evaluation the benefits of the AI/ML based attitude estimation.		The following figure shows the set up designed for evaluation the benefits of the AI/ML based attitude estimation.
	[[File:wp4-~~16_06~~.png\|100px\|frame\|center\|Experimentation set up for assessment of AI/ML based attitude estimation]]		[[File:wp4-16_07.png\|100px\|frame\|center\|Experimentation set up for assessment of AI/ML based attitude estimation]]

	The data feeding this assessment framework was a set of training&evaluation data synthetically generated. The idea was to design this dataset so it were fully representative of the diversity of inputs (baselines) provided to the estimator, and a wide diversity of errors applied to the baselines. For that, baslines where generated directly from a (uniformly distributed) random sample of attitude, plus		The data feeding this assessment framework was a set of training&evaluation data synthetically generated. The idea was to design this dataset so it were fully representative of the diversity of inputs (baselines) provided to the estimator, and a wide diversity of errors applied to the baselines. For that, baslines where generated directly from a (uniformly distributed) random sample of attitude, plus
	some (uniformly distributed) random error type. Several error types where described, depending on the number of baselines impacted and the magnitude of the error.		some (uniformly distributed) random error type. Several error types where described, depending on the number of baselines impacted and the magnitude of the error.
	[[File:wp4-~~16_07~~.png\|100px\|frame\|center\| Representativeness of the training and evaluation data in terms of attitude ranges covered, and errors in the baseline estimations.]]		[[File:wp4-16_08.png\|100px\|frame\|center\| Representativeness of the training and evaluation data in terms of attitude ranges covered, and errors in the baseline estimations.]]

Acorde: /* Validation */

2023-03-13T15:36:10Z

Validation

← Older revision		Revision as of 15:36, 13 March 2023
Line 79:		Line 79:

	However, this has not been the only mechanism for data generation. In order to serve to all the different research questions posed in C4D, ACORDE has also made a specific set of synthetic data generation, specifically for investigating the possible sustitution of the heurisctic functionality of the attitude estimation algorithm by a trained neural network.		However, this has not been the only mechanism for data generation. In order to serve to all the different research questions posed in C4D, ACORDE has also made a specific set of synthetic data generation, specifically for investigating the possible sustitution of the heurisctic functionality of the attitude estimation algorithm by a trained neural network.

			The following figure shows the set up designed for evaluation the benefits of the AI/ML based attitude estimation.
			[[File:wp4-16_06.png\|100px\|frame\|center\|Experimentation set up for assessment of AI/ML based attitude estimation]]

			The data feeding this assessment framework was a set of training&evaluation data synthetically generated. The idea was to design this dataset so it were fully representative of the diversity of inputs (baselines) provided to the estimator, and a wide diversity of errors applied to the baselines. For that, baslines where generated directly from a (uniformly distributed) random sample of attitude, plus
			some (uniformly distributed) random error type. Several error types where described, depending on the number of baselines impacted and the magnitude of the error.
			[[File:wp4-16_07.png\|100px\|frame\|center\| Representativeness of the training and evaluation data in terms of attitude ranges covered, and errors in the baseline estimations.]]

Acorde: /* Validation */

2023-03-13T15:25:43Z

Validation

← Older revision		Revision as of 15:25, 13 March 2023
Line 76:		Line 76:
	</gallery>		</gallery>

	A set of raw catured data, including multi-constellation (GPS-Galileo) observables wehere captured, which served for test the algorithm and even for further ~~refinementa~~ and evaluation.		A set of raw catured data, including multi-constellation (GPS-Galileo) observables wehere captured, which served for test the algorithm and even for further refinement and evaluation.

			However, this has not been the only mechanism for data generation. In order to serve to all the different research questions posed in C4D, ACORDE has also made a specific set of synthetic data generation, specifically for investigating the possible sustitution of the heurisctic functionality of the attitude estimation algorithm by a trained neural network.

Acorde: /* Improvements */

2023-03-13T15:10:12Z

Improvements

← Older revision		Revision as of 15:10, 13 March 2023
Line 54:		Line 54:
	</gallery>		</gallery>

	Beyond these graphically apparent example, ~~after~~ a systematic set of flights of different nature lead to 15%/23.3% average improvements in horizontal/vertical accuracy in static ~~condtions~~, and 11.9%/31.2 in dynamic conditions.		Beyond these graphically apparent example, a systematic analisys on a set of flights of different nature lead to 15%/23.3% average improvements for horizontal/vertical accuracy in static conditions, and 11.9%/31.2 in dynamic conditions.

	Another aspect addressed in the context of WP4 is an '''analysis to assess if state-of-the-art deep learning methodologies might provide feasible and advantageous solutions''' for embedded targets. Specifically, the possibility to enhance via AI/ML a currently heuristic functionality integrated '''on the multi-baseline attitude estimation algorithm of GLAD+''' was assessed (more details in D4.6 of C4D).		Another aspect addressed in the context of WP4 is an '''analysis to assess if state-of-the-art deep learning methodologies might provide feasible and advantageous solutions''' for embedded targets. Specifically, the possibility to enhance via AI/ML a currently heuristic functionality integrated '''on the multi-baseline attitude estimation algorithm of GLAD+''' was assessed (more details in D4.6 of C4D).

Acorde: /* Description */

2023-03-13T15:08:00Z

Description

← Older revision		Revision as of 15:08, 13 March 2023
Line 36:		Line 36:

	[[File:wp4-16_06.png\|frame\|center\|Basic architecture of the multi-gnss fusion sensor functionality of GLAD+]]		[[File:wp4-16_06.png\|frame\|center\|Basic architecture of the multi-gnss fusion sensor functionality of GLAD+]]


			==Improvements==

	COMP4DRONES has enable several improvements in the internals of this functionality.		COMP4DRONES has enable several improvements in the internals of this functionality.
Line 41:		Line 44:
	A former acievement was a '''configurable algorithm that can adapt to different types of low-cost receivers'''. This is required, as a lesson learned from C4D is that the types of preprocessins (e.g., smoothing, bias adaptations) supported by different low-cost GNSS receiver brands can be very different, which these aspects will have great error impact on the position/attitude estimation unless not considered.		A former acievement was a '''configurable algorithm that can adapt to different types of low-cost receivers'''. This is required, as a lesson learned from C4D is that the types of preprocessins (e.g., smoothing, bias adaptations) supported by different low-cost GNSS receiver brands can be very different, which these aspects will have great error impact on the position/attitude estimation unless not considered.

	A prominent enhancement of the navigation software has been the exploitation of the novel low-cost receivers integrated in the HW platform developed in COMP4DRONES. The new receivers enable '''low-cost multi-constellation data, which need to be integrated and exploited in the navigation algorithm'''. This has been mostly validated for the standard positioning (trilateration) and more specifically, for the TI block of the GLAD+ architecture (more details on D4.3).		A prominent enhancement of the navigation software has been the exploitation of the novel low-cost receivers integrated in the HW platform developed in COMP4DRONES. The new receivers enable '''low-cost multi-constellation data, which need to be integrated and exploited in the navigation algorithm'''. This has been mostly validated for the standard positioning (trilateration) and more specifically, for the TI block of the GLAD+ architecture (more details on D4.3). These receivers also enable specific reactive security features, an aspect also addressed in C4D and further explained in [[WP5-11_ACO]].

	These receivers also enable specific reactive security features, an aspect also addressed in C4D and further explained in [[WP5-11_ACO]].

	Another aspect addressed in the context of WP4 is an '''analysis to assess if state-of-the-art deep learning methodologies might provide feasible and advantageous solutions''' for embedded targets. Specifically, the possibility to enhance via AI/ML a currently heuristic functionality integrated '''on the multi-baseline attitude estimation algorithm of GLAD+''' was assessed (more details in D4.6 of C4D).

	~~Last, but not least, in the framework of WP4, ACORDE has also provided a "Mavlink interface" to ease SW, and therefore GLAD+ system integration (more details in D4.8).~~


	~~Some of~~ the ~~reported improvements were~~ on the accuracy of the low-cost geo-referenced position.		The following figures graphically show the '''improvement on the accuracy''' of the low-cost geo-referenced position.
	The following figures show GLAD+ position output (blue) vs RTK-postprocesed output (green) in one flight test performed in COMP4DRONES in a case where GLAD+ is configured to use only GPS, and in another case where GLAD+ is confiured to use multiconstellation (both GPS and Galileo).		The following figures show GLAD+ position output (blue) vs RTK-postprocesed output (green) in one flight test performed in COMP4DRONES in a case where GLAD+ is configured to use only GPS, and in another case where GLAD+ is confiured to use multiconstellation (both GPS and Galileo).

	<gallery widths=500px heights=500px caption="Accuracy improvement with Multi-Constellation (GPS-Galileo) in one flight test performed in COMP4DRONES">>		<gallery widths=500px heights=500px caption="Accuracy improvement with Multi-Constellation (GPS-Galileo) in one flight test performed in COMP4DRONES">>

	File:wp4-16-01.gif\|center\|Only GPS		File:wp4-16-01.gif\|center\|Only GPS
	File:wp4-16-02.gif\|center\|Multiconstellation (GPS-GAL)		File:wp4-16-02.gif\|center\|Multiconstellation (GPS-GAL)
			</gallery>

	</~~gallery>~~		Beyond these graphically apparent example, after a systematic set of flights of different nature lead to 15%/23.3% average improvements in horizontal/vertical accuracy in static condtions, and 11.9%/31.2 in dynamic conditions.

			Another aspect addressed in the context of WP4 is an '''analysis to assess if state-of-the-art deep learning methodologies might provide feasible and advantageous solutions''' for embedded targets. Specifically, the possibility to enhance via AI/ML a currently heuristic functionality integrated '''on the multi-baseline attitude estimation algorithm of GLAD+''' was assessed (more details in D4.6 of C4D).

	~~Beyond these graphically apparent example~~, ~~after a systematic set~~ of ~~flights of different nature~~, ACORDE ~~reported~~		Last, but not least, in the framework of WP4, ACORDE has also provided a "Mavlink interface" to ease SW, and therefore GLAD+ system integration (more details in D4.8).

	==Validation==		==Validation==

Acorde: /* Improvements */

2023-03-13T15:03:06Z

Improvements

← Older revision		Revision as of 15:03, 13 March 2023
Line 49:		Line 49:
	Last, but not least, in the framework of WP4, ACORDE has also provided a "Mavlink interface" to ease SW, and therefore GLAD+ system integration (more details in D4.8).		Last, but not least, in the framework of WP4, ACORDE has also provided a "Mavlink interface" to ease SW, and therefore GLAD+ system integration (more details in D4.8).

	~~==Improvements==~~

	Some of the reported improvements were on the accuracy of the low-cost geo-referenced position.		Some of the reported improvements were on the accuracy of the low-cost geo-referenced position.