This paper discusses implementation and use of robotic fish in science education. Biologically inspired robotic fish have been found to be particularly engaging, likely due to the additional connections to the natural world they can offer. Thus, a few robotic fish exhibits have been deployed to engage and educate visitors in public aquariums and expositions. However, such exhibits are often limited in the level of interactivity they afford, which is known to be a key factor in informal science education. A novel interactive robotic fish exhibit has been developed at the Dynamical Systems Laboratory of New York University Polytechnic School of Engineering. The platform has been well received by visitors and has contributed to increasing their interest in robotics. Integrating natural user interfaces and autonomous charging into the exhibit are expected to further enhance the visitor experience and strengthen the feasibility of robotics-based informal science education {AQ: An edit has been made in this sentence “Integrating natural… science education.”, for the verb to match with its subjects. Please check and correct if necessary.}.

Article

Informal science education is the process of scientific learning that takes place outside of the classrooms and academic institutions [1, 2]. It is the most predominant form of learning across lifelong education, is spontaneous in nature, and has practically unlimited opportunities [1, 2]. Informal learning can occur through visits to museums and galleries, participation in science festivals, and even watching educational programs [1].

For visitors to informal science venues, robotics has been shown to be an effective tool to elicit their interest, as it often affords several elements of novelty [3]. Further, robotics offers quick feedback for participants to test new ideas or reinforce preexisting knowledge [4, 5]. Thus, a number of robotics-based exhibits, such as the exploratory rover [6], robotic dolphin [7], and remotely- controlled miniature boats [8], have been designed to increase visitors' interest in robotics, while delivering important topics in science, like space exploration [6] and environmental mapping [8].

Biologically-inspired robotic fish have been found to be particularly engaging [7, 9, 10], likely due to the additional connections to the natural world they can offer [11]. Thus, a few robotic fish exhibits have been deployed to engage and educate visitors in public aquariums and expositions [9, 10]. However, such exhibits are often limited in the level of interactivity they afford, which is known to be a key factor in informal science education [12, [13].

A novel interactive robotic fish exhibit has been developed at the Dynamical Systems Laboratory of New York University (NYU) Polytechnic School of Engineering to address this gap [14]. The exhibit features Commodore (Figure 1), a robotic fish based on a multi-link design with a pitch and buoyancy control system for three-dimensional biologically-inspired swimming. An ad hoc iDevice application with three control modes varying in the level of robot autonomy provides visitors with a unique experience to control the robotic fish swimming.

graphic

Biologically-Inspired Robotic Fish

Commodore is designed with inspiration from the Atlantic scup fish, Stenotomus chrysops. The cover of the robotic fish is fabricated through a rapid prototyping machine from solid-packing acrylonitrile butadiene styrene plastic. Its mechanical design includes a three degree-of- freedom motorized tail, electronics housing, and pitch and buoyancy control systems (Figure 2). The undulation of the tail enables the robotic fish to swim in two dimensions, and the pitch and buoyancy control systems allow the fish to dive and surface. The robotic fish measures 48 cm in length, 19 cm in height, and 10 cm in width.

graphic

Electronics

The primary electronic components of Commodore include an Arduino-based microcontroller, a radio transceiver, and a 2200 mAh 7.4V lithium polymer battery. A battery charge sensor, composed of a voltage divider, is incorporated to assist with monitoring the battery power level. Two infrared (IR) sensors, facing 120° apart, are mounted on the front of the robotic fish to provide navigational information when walls or obstacles are detected.

Pitch and Buoyancy Control

The combination of the pitch and buoyancy control systems is used to adjust the depth of the robotic fish. Specifically, the buoyancy control system heuristically sets the robotic fish to neutral buoyancy by adjusting the water content within a rigid volume tank.

The pitch angle is regulated by shifting a balance mass with a 2.2 cm travel length. This mass is used to control the center of gravity of Commodore with respect to the center of buoyancy, thus tilting it upwards (positive pitch) or downwards (negative pitch). When Commodore is close to neutral buoyancy, adjusting the pitch angle while beating the tail allows rising and diving.

The buoyancy system comprises a rigid enclosure, fabricated using rapid prototyping, which contains two separate chambers (Figure 2). One chamber encases two motors, pumping water between the second holding chamber and the environment through airline tubing, while the other chamber simply holds the water. Commodore can reach neutral buoyancy as the water in the latter holding chamber is varied.

Fish Tail

The tail of the robotic fish consists of three servomotors connected by rigid plastic links and a rubber silicone caudal fin. Coordinating these three motors causes the tail to undulate and the Commodore to swim. A simplified model for carangiform fish swimming [15] is used to mimic the swimming pattern of a real scup fish.

Communication and Idevice Application

A communication network is developed to operate Commodore from a custom iDevice application (Figure 3). In the network, a user datagram protocol (UDP) is utilized to communicate between the base station, a desktop computer, and the iDevice for delivering important messages, such as the control mode and its key parameters. The base station comprises a microcontroller, a radio shield, an Ethernet shield, and a wireless router, and is used to translate commands from UDP to radio signals for Commodore.

graphic

The three control modes have separate screens on the iDevice application (Figure 4). i) In the manual mode, the user has the option to use sliders and buttons to directly control the robotic fish, including the tail-beat frequency, steering, diving, and surfacing. ii) In the semi-autonomous mode, the user can generate a waypoint path for Commodore to follow, by tapping on the iDevice screen that shows a real-time bird’s-eye view of the tank through an overhead webcam. The robotic fish then automatically swims along the prescribed path by implementing a proportional-integral- derivative (PID) controller on its orientation, using feedback from the webcam. iii) In the autonomous mode, an animated fish is displayed on the screen, offering the user a visual representation of the swimming decisions of Commodore. When one of the IR sensors detects an object within its vicinity, the user observes a red brick wall appearing in front of the animated fish, and the robotic fish and its animation both react to turn away from the obstacle.

graphic

Applications in Informal Science

The robotic fish platform was deployed in various informal science events and venues, in April 2014 at the USA Science and Engineering Festival, in May 2014 at NYU Polytechnic School of Engineering Research Expo, and in June 2014 at the World Science Festival (Figure 5). All events were open and accessible to the general public, whereby thousands of visitors interacted with Commodore over the course of the summer of 2014.

graphic

At these events, data from 224 visitors were recorded: users spent on average 85 seconds in the manual mode, 69 seconds in the semi-autonomous mode, and 5 seconds in the autonomous mode. Due to the nature of the events, the popularity of the exhibit amongst youngsters, and the high visitor traffic, when opportunity arose to use the iDevice application, visitors were inclined to select an interactive control mode over the autonomous, passive mode.

An optional questionnaire was administered to visitors after controlling the robotic fish, to acquire feedback on the usability of the platform and learn their thoughts on pertinent environmental issues. Users indicated that the experience was enjoyable, rating the exhibit on average 4.6 out of 5 stars, and identified marine pollution, overfishing, and climate change as the most pressing environmental issues we are currently facing.

To allow visitors additional time to interact with the platform and to examine uninterrupted interactions with Commodore, the platform was customized to operate in a 1000 liter tank (Figure 6) and demonstrated to the public in New York City, at The River Project from August 19, 2014 to September 22, 2014 and at the Brooklyn Children’s Museum from October 9, 2014 to January 19, 2015. In this platform, the robotic fish was presented alongside educational material that was created based on the usability studies and interactions with education professionals.

graphic

Outlook on Future Research

To further expand on the degree of interactivity of Commodore, we are exploring natural user interfaces. Such interfaces have been shown to be effective tools for enhancing user engagement and learning [16-[18]. A proposed platform to control a robotic fish with a natural user interface is illustrated in Figure 7. This system can possibly involve a minigame, such as to collect water temperature, to further encourage user engagement [19].

graphic

In a parallel research effort, we are also investigating a new autonomous charging station for the robotic fish to minimize maintenance by museum personnel. A proposed platform for enabling continuous use of an untethered robotic fish through a direct contact approach is sketched in Figure 8.

graphic

Conclusions

An interactive robotic fish exhibit was developed to aid informal science learning. The platform has been well received by visitors and has contributed to increasing their interest in robotics. Integrating natural user interfaces and autonomous charging into the exhibit is expected to further enhance the visitor experience and strengthen the feasibility of robotics-based informal science education.

Acknowledgment

The authors would like to thank Drs. Mary Leou and Catherine Milne for Important feedback on the informal science education aspects of the platform; Raymond LeGrand and Chengxing Zhang for the implementation of iDevice applications; Jayhwan Cheong, Mauro DeBellis, Chengxing Zhang, and Chun Hsien-Wu for assistance with hardware development; Jeffrey Laut and Tiziana Bartolini for help with planning and implementing the events; The River Project and the Brooklyn Children's Museum for graciously hosting Commodore; and all the visitors for their valuable feedback. This research was supported by the National Science Foundation under grant nos. CMMI-0745753, DGE-0741714, and DRL-1200911.

References

References
1.
Ainsworth
H. L.
,
Eaton
S. E.
,
Formal, Non-formal and Informal Learning in the Sciences
,
Calgary, Canada
:
Onate Press
,
2010
.
2.
Banks
J.
,
Au
K.
,
Ball
A.
,
Bell
P.
,
Gordon
E.
,
Gutiérrez
K.
, et al. ., “Learning in and out of school in diverse environments: Life-long, life-wide, life-deep,” The LIFE Center, Seattle, WA2007.
3.
Falk
J. H.
,
Balling
J. D.
,
“The field trip milieu: Learning and behavior as a function of contextual events”
,
J. Educ. Res.
, Vol.
76
, no.
1
, pp.
22
28
,
1982
.
4.
Wilkinson
C.
,
Bultitude
K.
,
Dawson
E.
,
“Oh yes, robots! People like robots; the robot people should do something: perspectives and prospects in public engagement with robotics”
,
Sci. Commun.
, vol.
33
, no.
3
, pp.
367
397
,
2010
.
5.
Miller
D. P.
,
Nourbakhsh
I. R.
,
Siegwart
R.
, “Robots for education”, in
Springer Handbook of Robotics
, ed
Berlin, Germany
:
Springer
,
2008
, pp.
1283
1301
.
6.
Nourbakhsh
I.
,
Hamner
E.
,
Ayoob
E.
,
Porter
E.
,
Dunlavey
B.
,
Bernstein
D.
, et al. .,
“The Personal Exploration Rover: Educational assessment of a robotic exhibit for informal learning venues”
,
Int. J. Eng. Educ.
, vol.
22
, no.
4
, pp. ,
777
791
,
2006
.
7.
Yu
J.
,
Wang
M.
,
Tan
M.
, and
Zhang
J.
,
“Three-dimensional swimming”
,
IEEE Robot. Automat. Mag.,
vol.
18
, no.
4
, pp.
47
58
,
2011
.
8.
Laut
J.
,
Wright
S.
,
Nov
O.
, and
Porfiri
M.
,
“Gowanus Voyage: Where mechatronics, public art, community members, and environmental science meet”
,
IEEE Control Syst.,
vol.
34
, no.
1
, pp.
60
64
,
2014.
9.
Terada
Y.
, and
Yamamoto
I.
,
“An animatronic system including lifelike robotic fish”
,
Proc. IEEE,
vol.
92
, no.
11
, pp.
1814
1820
,
2004.
10.
Hu
H.
,
“Biologically inspired design of autonomous robotic fish at Essex”
, in
Proc. IEEE SMC UK-RI Chapter Conf.,
Sheffield, UK,
2006
, pp. 3–8.
11.
Falk
J. H.
,
Reinhard
E. M.
,
Vernon
C.
,
Bron-nenkant
K.
,
Heimlich
J. E.
, and
Deans
N. L.
,
Whyzoos & aquariums matter: Assessing the impact of a visit to a zoo or aquarium
.
Silver Spring, MD: Association of Zoos & Aquariums
,
2007
.
12.
Allen
S.
,
“Designs for learning: Studying science museum exhibits that do more than entertain”
,
Sci. Educ
, vol.
88
, no.
S1
, pp.
S17
S33
,
2004
.
13.
Schneider
B.
, and
Cheslock
N.
,
Measuring Results: Gaining Insight on Behavior Change Strategies and Evaluation Methods from Environmental Education, Museum, Health, and Social Marketing Programs
.
San Francisco, CA
:
Coevolution Institute
,
2003
.
14.
Phamduy
P.
,
LeGrand
R.
, and
Porfiri
M.
,
“Design and characterization of an interactive iDevice-controlled robotic fish for informal science education”
,
IEEE Robot. Automat. Mag.,
accepted for publication,
2015
.
15.
Barrett
D. S.
,
Triantafyllou
M. S.
,
Yue
D. K. P.
,
Grosenbaugh
M. A.
, and
Wolfgang
M. J.
,
“Drag reduction in fish-like locomotion”
,
J. Fluid Mech
, vol.
392
, no.
1
, pp.
183
212
,
1999
.
16.
Hsu
H. M. J.
,
“The potential of Kinect in education”
,
Int. J. Inform. Educ. Technol.
, vol.
1
, no.
5
, pp.
365
370
,
2011
.
17.
Alisi
T. M.
,
Del Bimbo
A.
, and
Valli
A.
,
“Natural interfaces to enhance visitors’ experiences”
,
IEEE Multimedia
, vol.
12
, no.
3
, pp.
80
85
,
2005
.
18.
Johnson
K.
,
Pavleas
J.
, and
Chang
J.
,
“Kinecting to mathematics through embodied interactions”
,
Computer
, vol.
46
, no.
10
, pp.
101
104
,
2013
.
19.
Lange
B.
,
Chang
C. Y.
,
Suma
E.
,
Newman
B.
,
Rizzo
A. S.
, and
Bolas
M.
, “Development and evaluation of low cost game-based balance rehabilitation tool using the Microsoft Kinect sensor”, in Proc. Ann. Int. Conf. IEEE Eng. Med. Biol. Soc., Boston, MA, USA,
2011
.