Can Small Scale Search Behaviours Enhance Large-Scale Navigation?
341
position, the other close to the beacon. The Cyclops converges towards the beacon in most trials, with only a low percentage running away. Due to the sinuous
movements, a large area is covered. Almost all endpoints are located between
the starting position and the beacon, therefore it is more successful than BB.
The Braitenberg-CPG version converges towards the beacon in each trial. If it
does not detect the beacon initially, the CPG driven small scale search steers the
agent around (driven by Type III activity), until the stimulus becomes available
and the proportional steering mode takes over (mediated by Type I neurons).
The endpoints exhibit a smaller mean distance to the goal than in the other
conditions.
4 Discussion
We have developed a spiking neural network model of an insect-inspired CPG
which is used to underpin the steering behaviour of a Braitenberg-like vehicle.
We show that a Braitenberg-only setup, or familiarity modulated CPG, have
limited success navigating towards a beacon, while combining these approaches
increases success. In future the system will be explored in terms of its biological plausibility to better understand the LAL network and it connection to
other brain regions [10] and in more realistic scenarios we will investigate if this
framework can improve navigation algorithms.
References
1. Wilson, D.: The central nervous control of locust flight. J. Exp. Biol. 38(471–490),
8 (1961)
2. Ijspeert, A.J.: Central pattern generators for locomotion control in animals and
robots: a review. Neural Netw. 21, 642–653 (2008)
3. Grillner, S., Wallen, P., Brodin, L.: Neuronal network generating locomotor
behaviour in lamprey: circuitry, transmitters, membrane properties and simulation. Ann. Rev. Neurosci. 14, 169–99 (1991)
4. Selverston, A.I.: Invertebrate central pattern generator circuits. Philos. Trans. R.
Soc. 365, 2329–2349 (2010)
5. Kanzaki, R., Sugi, N., Shibuya, T.: Self-generated zigzag turning of Bombyx-Mori
males during pheromone-mediated upwind walking. Zool. Sci. 9, 515–527 (1992)
6. Zorovic, M., Hedwig, B.: Descending brain neurons in the cricket Gryllus bimaculatus (de Geer): auditory responses and impact on walking. J. Comp. Physiol.
Neuroethol. Sens. Neural Behav. Physiol. 199, 25–34 (2013)
7. Chiba, R., et al.: Neural network estimation of LAL/VPC resions of silkmoth using
genetic algorithm. In: The 2010 IEEE/RSJ International Conference on Intelligent
Robots and Systems (2010)
8. Steinbeck, F., Adden, A., Graham, P.: Connecting brain to behaviour: a role for
general purpose steering circuits in insect orientation? J. Exp. Biol. 223 (2020)
9. Zeil, J., Hofmann, M.I., Chahl, J.S.: Catchment areas of panoramic snapshots in
outdoor scenes. J. Opt. Soc. Am. A 20(3), 450–469 (2003)
10. Sun, X., Yue, S., Mangan, M.: A decentralised neural model explaining optimal integration of navigational strategies in insects. eLife 2020;9:e54026 (2020).
https://doi.org/10.7554/eLife.54026
341
position, the other close to the beacon. The Cyclops converges towards the beacon in most trials, with only a low percentage running away. Due to the sinuous
movements, a large area is covered. Almost all endpoints are located between
the starting position and the beacon, therefore it is more successful than BB.
The Braitenberg-CPG version converges towards the beacon in each trial. If it
does not detect the beacon initially, the CPG driven small scale search steers the
agent around (driven by Type III activity), until the stimulus becomes available
and the proportional steering mode takes over (mediated by Type I neurons).
The endpoints exhibit a smaller mean distance to the goal than in the other
conditions.
4 Discussion
We have developed a spiking neural network model of an insect-inspired CPG
which is used to underpin the steering behaviour of a Braitenberg-like vehicle.
We show that a Braitenberg-only setup, or familiarity modulated CPG, have
limited success navigating towards a beacon, while combining these approaches
increases success. In future the system will be explored in terms of its biological plausibility to better understand the LAL network and it connection to
other brain regions [10] and in more realistic scenarios we will investigate if this
framework can improve navigation algorithms.
References
1. Wilson, D.: The central nervous control of locust flight. J. Exp. Biol. 38(471–490),
8 (1961)
2. Ijspeert, A.J.: Central pattern generators for locomotion control in animals and
robots: a review. Neural Netw. 21, 642–653 (2008)
3. Grillner, S., Wallen, P., Brodin, L.: Neuronal network generating locomotor
behaviour in lamprey: circuitry, transmitters, membrane properties and simulation. Ann. Rev. Neurosci. 14, 169–99 (1991)
4. Selverston, A.I.: Invertebrate central pattern generator circuits. Philos. Trans. R.
Soc. 365, 2329–2349 (2010)
5. Kanzaki, R., Sugi, N., Shibuya, T.: Self-generated zigzag turning of Bombyx-Mori
males during pheromone-mediated upwind walking. Zool. Sci. 9, 515–527 (1992)
6. Zorovic, M., Hedwig, B.: Descending brain neurons in the cricket Gryllus bimaculatus (de Geer): auditory responses and impact on walking. J. Comp. Physiol.
Neuroethol. Sens. Neural Behav. Physiol. 199, 25–34 (2013)
7. Chiba, R., et al.: Neural network estimation of LAL/VPC resions of silkmoth using
genetic algorithm. In: The 2010 IEEE/RSJ International Conference on Intelligent
Robots and Systems (2010)
8. Steinbeck, F., Adden, A., Graham, P.: Connecting brain to behaviour: a role for
general purpose steering circuits in insect orientation? J. Exp. Biol. 223 (2020)
9. Zeil, J., Hofmann, M.I., Chahl, J.S.: Catchment areas of panoramic snapshots in
outdoor scenes. J. Opt. Soc. Am. A 20(3), 450–469 (2003)
10. Sun, X., Yue, S., Mangan, M.: A decentralised neural model explaining optimal integration of navigational strategies in insects. eLife 2020;9:e54026 (2020).
https://doi.org/10.7554/eLife.54026