Skip to main content

About this book

This book presents the result of a joint effort from different European Institutions within the framework of the EU funded project called SPARK II, devoted to device an insect brain computational model, useful to be embedded into autonomous robotic agents.
Part I reports the biological background on Drosophila melanogaster with particular attention to the main centers which are used as building blocks for the implementation of the insect brain computational model.
Part II reports the mathematical approach to model the Central Pattern Generator used for the gait generation in a six-legged robot. Also the Reaction-diffusion principles in non-linear lattices are exploited to develop a compact internal representation of a dynamically changing environment for behavioral planning.
In Part III a software/hardware framework, developed to integrate the insect brain computational model in a simulated/real robotic platform, is illustrated. The different robots used for the experiments are also described. Moreover the problems related to the vision system were addressed proposing robust solutions for object identification and feature extraction.
Part IV includes the relevant scenarios used in the experiments to test the capabilities of the insect brain-inspired architecture taking as comparison the biological case. Experimental results are finally reported, whose multimedia can be found in the SPARK II web page:

Table of Contents


Models of the Insect Brain: From Neurobiology to Computational Intelligence


Chapter 1. Neurobiological Models of the Central Complex and the Mushroom Bodies

This study reviews the actual knowledge on functions of the central complex (CX) and the mushroom bodies (MBs) in a genetic model insect, the fly

Drosophila melanogaster

. Ongoing research of UNIMAINZ and respective data are included. Reference is made to other insects, where respective functions are not yet studied in


. Neuroanatomical information is reported with regard to the general flow of information in these central brain neuropils. Particular projection systems and circuits are taken into account where this can be linked to functions. Models are developed.

R. Strauss

Chapter 2. A Computational Model for the Insect Brain

As seen in the Chap.


, the fruit fly

Drosophila melanogaster

is an extremely interesting insect because it shows a wealth of complex behaviors, despite its small brain. Nowadays genetic techniques allow to knock out the function of defined parts or genes in the


brain. Together with specific mutants which show similar defects in those parts or genes, hypothesis about the functions of every single brain part can be drawn. Based upon the results reported in the Chap.


, a computational model of the fly


has been designed and implemented to emulate the functionalities of the two relevant centres present in insects: the Mushroom Bodies and the Central Complex. Their actions and inter-actions are adapted from the neurobiological prospective to a computational implementation. A complete block scheme is proposed where the proved or conjectured interactions among the identified blocks are depicted. Several simulations results are finally provided to demonstrate the capability of the system both considering specific parts of the complete structure for comparison with insect experiments, and the whole model for more complex simulations.

P. Arena, L. Patanè, P. S. Termini

Complex Dynamics for Internal Representation and Locomotion Control


Chapter 3. Compact Internal Representation of Dynamic Environments: Simple Memory Structures for Complex Situations

In this chapter the novel concept of Compact Internal Representation (CIR) is introduced as a generalization of the internal representation extensively used in literature as a base for cognition and consciousness. CIR is suitable to represent dynamic environments and their potential interactions with the agent as static (time-independent) structures, suitable to be stored, managed, compared and recovered by memory. In this work the application of CIR as the cognitive core of moving autonomous artificial agents is presented in the context of collision avoidance against dynamical obstacles. The structure that emerges, even if not directly related to the insect neurobiology, is quite simple and could enhance the capabilities of the computational model already presented in the previous chapter, in view of its robotic implementation.

J. A. Villacorta-Atienza, M. G. Velarde, V. A. Makarov

Chapter 4. CPG for Motor Control

In this Chapter the main results related to locomotion control are reported. Theoretical results referring to Contraction theory are exploited to design reaction diffusion dynamical systems able to work as stable Central Pattern Generators for multipodal robotic structures. A series of strategies are also discussed referring to the control of migration through different locomotion patterns as well as to the steering control for trajectory planning. Relevant parameters are also outlined in view of their modulation for a low low-level feedback control. Both theoretical and experimental results are reported to verify the suitability of the approach.

E. Arena, P. Arena, L. Patanè

Chapter 5. A Prototype 2N-Legged (insect-like) Robot. A Non-Linear Dynamical System Approach

A nonlinear closed lattice or ring is proposed as a central pattern generator (CPG) for controlling hexapodal robots. We show that the ring composed of six anharmonically interacting units coupled to the limb actuators permits to reproduce typical hexapod gaits. We provide an electronic circuit implementation of the CPG providing the corresponding gaits. Then we propose a method to incorporate the actuator (motor) and leg dynamics in the units of the CPG. With this electro-mechanical device we close the loop CPG—environment—CPG, thus obtaining a decentralized approach for the leg control that does not require higher level CPG intervention during locomotion in a non-smooth hence non flat landscape. The gaits generated by our CPG are not rigid, but adapt to obstacles faced by the robot.

E. del Rio, M. G. Velarde

Software/Hardware Cognitive Architectures


Chapter 6. A Robotic Simulation Framework for Cognitive Systems

The insect brain computational model introduced in Part I of the book was there demonstrated through simple simulation results which showed the performance of the main blocks involved. In this chapter a dedicated software/hardware framework for cooperative and bio-inspired cognitive architectures, were the brain computational model was embedded, is presented. Here a complete description of the system, named Robotic Simulations for Cognitive Systems (RS4CS), will be introduced in order to show potentialities and capabilities. Moreover, the design choices and implementation issues related to the proposed robotic programming environment will be here addressed. The framework can be interfaced with robot prototypes mediating the sensory motor loop or with 2D and 3D kinematic and dynamic simulation environments.

P. Arena, L. Patanè, A. Vitanza

Chapter 7. Robotic Platforms

To evaluate the capabilities of the insect brain model different robotic platforms have been considered. The different blocks of the cognitive architecture, inspired by MBs and CX, can be used as control systems both for legged and wheeled robots. This chapter reports the characteristics of the robotic platforms including information on the mechanical structure, sensory system, software and hardware low level control architecture. In particular the wheeled robot considered is the Pioneer P3AT, a commercial platform for indoor and outdoor applications, suitably modified to host new sensors and control boards. Concerning legged and hybrid robots, a series of robots have been developed to exploit the insect brain main functions on different platforms.

I. Aleo, P. Arena, S. De Fiore, L. Patanè, M. Pollino, C. Ventura

Chapter 8. Compact Internal Representation of Dynamic Environments: Implementation on FPGA

Animals for surviving have developed cognitive abilities allowing them an abstract representation of the environment. This internal representation (IR) may contain a huge amount of information concerning the evolution and interactions of the animal and its surroundings. The temporal information is needed for Internal Representations of dynamic environments and is one of the most subtle points in its implementation as the information needed to generate the IR may eventually increase dramatically. Chapter


in the book proposed the compression of the spatio-temporal information into only space, leading to a stable structure suitable to be the base for complex cognitive processes in what has been called Compact Internal Representation (CIR). The Compact Internal Representation is especially suited to be implemented in autonomous robots as it provides global strategies for the interaction with real environments. This chapter describes an FPGA implementation of a Causal Neural Network based on a modified FitzHugh-Nagumo neuron to generate a Compact Internal Representation of dynamic environments for roving robots to avoid dynamic and static obstacles.

L. Salas-Paracuellos, L. Alba-Soto

Chapter 9. Visual Routines for Cognitive Systems on the Eye-RIS Platform

The purpose of this chapter is to describe how different visual routines can be developed and embedded in the AnaFocus’ Eye-RIS Vision System on Chip (VSoC) to close the perception to action loop within the roving robots developed for the testing of the insect brain models. The Eye-RIS VSoC employs a bio-inspired architecture where image acquisition and processing are intermingled and the processing itself is carried out in two steps. At the first step, processing is fully parallel owing to the concourse of dedicated circuit structures which are integrated close to the sensors. At the second step, processing is realized on digitally-coded information data by means of digital processors. The Eye-RIS VSoC systems provide with image-processing capabilities and speed comparable to high-end conventional vision systems without the need for high-density image memory and intensive digital processing. Current perceptual schemes are often based on information derived from visual routines. Since real world images are quite complex to be processed for perceptual needs with traditional approaches, more computationally feasible algorithms are required to extract the desired features from the scene in real time, to efficiently proceed with the consequent action. In this chapter the development of such algorithms and their implementation taking full advantage of the sensing-processing capabilities of the Eye-RIS VSoC are described.

D. J. Caballero-Garcia, A. Jimenez-Marrufo

Scenarios and Experiments


Chapter 10. Experimental Scenarios

This Chapter describes a number of different scenarios and related experiments envisaged according to the various design principles for cognitive architectures. The primary approach taken into consideration was the biologically driven one, so the scenarios have been prepared at the aim to demonstrate how efficiently the insect brain computational model built succeeds in reproducing and enhancing the insect capabilities already addressed in Neurobiology.

P. Arena, L. Patanè

Chapter 11. Robotic Experiments and Comparisons

This Chapter will present the final results obtained in testing the insect brain computational model. A series of experiments have been assessed in order to evaluate the capabilities of the proposed control architecture, in comparison with the behaviour shown by real flies while performing the same tasks.

P. Arena, S. De Fiore, L. Patanè, P. S. Termini, A. Vitanza
Additional information