An Interface Platform for Robotic Neuromorphic Systems
Abstract
:1. Introduction
2. Interface Board Design and Specification
2.1. SpiNNaker
2.2. Arduino Due
2.3. Hardware Links
3. Communication Protocol
3.1. 2-of-7 Coding
3.2. 2-Phase Handshake Protocol
3.3. Encoding/Decoding Packets
4. Evaluation
4.1. Network Topology
4.2. Up-Link Data Rate
4.3. Down-Link Data Rate
4.4. Accuracy
4.5. Latency
5. Discussion
6. Conclusions
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Conflicts of Interest
Abbreviations
ACK | Acknowledge Packet |
ARM | Advanced RISC Machine Processor |
ATP | Advanced Processor Technologies Research Group |
CPLD | Complex Programmable Logic Device |
DAC | Digital-to-Analog Converter |
DC | Direct Current |
DVS | Dynamic Video Sensor |
EOP | End of Procedure |
FPGA | Field Programmable Gate Arrays |
GPIO | General Purpose Input Output |
IB | Interface Board |
MC | Multi Cast |
MCU | Micro Controller Unit |
NN | Nearest Neighbour |
NRZ | Non Return to Zero |
PCB | Printed Circuit Board |
SNN | Spiking Neural Network |
UART | Universal Asynchronous Receiver-Transmitter |
USB | Universal Serial BUS |
References
- Ponulak, F.; Kasiński, A. Supervised learning in spiking neural networks with ReSuMe: Sequence learning, classification, and spike shifting. Neural Comput. 2010, 22, 467–510. [Google Scholar] [CrossRef] [PubMed]
- Ghosh-Dastidar, S.; Adeli, H. Spiking neural networks. Int. J. Neur. Syst. 2009, 19, 295–308. [Google Scholar] [CrossRef] [PubMed]
- Wei, D.; Harris, J.G. Signal reconstruction from spiking neuron models. In Proceedings of the 2004 IEEE International Symposium on Circuits and Systems (IEEE Cat. No.04CH37512), Vancouver, BC, Canada, 23–26 May 2004. [Google Scholar]
- Furber, B.S.; Galluppi, F.; Temple, S.; Plana, L.A. The SpiNNaker project. Proc. IEEE 2014, 102, 652–665. [Google Scholar] [CrossRef]
- Davies, S.; Patterson, C.; Galluppi, F.; Rast, A.D.; Lester, D.; Furber, B.S. Interfacing real-time spiking i/o with the spinnaker neuromimetic architecture. Aust. J. Intell. Inf. Process. Syst. 2010, 11, 7–11. [Google Scholar]
- iniLabs. Pushbot. Available online: https://inilabs.com/products/pushbot/ (accessed on 1 May 2021).
- Galluppi, F.; Denk, C.; Meiner, M.C.; Stewart, T.C.; Plana, L.A.; Eliasmith, C.; Furber, S.; Conradt, J. Event-based neural computing on an autonomous mobile platform. In Proceedings of the 2014 IEEE International Conference on Robotics and Automation (ICRA), Hong Kong, China, 31 May–7 June 2014; pp. 2862–2867. [Google Scholar]
- Bekolay, T.; Bergstra, J.; Hunsberger, E.; DeWolf, T.; Stewart, T.C.; Rasmussen, D.; Choo, X.; Voelker, A.R.; Eliasmith, C. Nengo: A Python tool for building large-scale functional brain models. Front. Neuroinform. 2013, 3, 48. [Google Scholar] [CrossRef] [PubMed]
- Plana, L.A. Interfacing Aer Devices to Spinnaker Using an Fpga. Available online: http://spinnakermanchester.github.io/docs/spinn-app-8.pdf (accessed on 1 July 2021).
- Denk, C.; Llobet-Blandino, F.; Galluppi, F.; Plana, L.A.; Furber, S.; Conradt, J. Real-time interface board for closed-loop robotic tasks on the spinnaker neural computing system. In Proceedings of the International Conference on Artificial Neural Networks (ICANN), Sofia, Bulgaria, 10–13 September 2013; pp. 467–474. [Google Scholar]
- Cheng, R.; Mirza, K.B.; Nikolic, K. Neuromorphic robotic platform with visual input, processor and actuator, based on spiking neural networks. Appl. Syst. Innov. 2020, 3, 28. [Google Scholar] [CrossRef]
- Furber, S.; Bogdan, P. SpiNNaker: A Spiking Neural Network Architecture; Now Publishers: Delft, The Netherlands, 2020. [Google Scholar]
- Furber, S.; Temple, S. Neural systems engineering. J. R. Soc. Interface 2006, 4, 193–206. [Google Scholar] [CrossRef] [PubMed]
- Furber, S.B.; Temple, S.; Brown, A.D. High-performance computing for systems of spiking neurons. In Proceedings of the AISB’06 Workshop on GC5: Architecture of Brain and Mind, Bristol, UK, 3–4 April 2006; Volume 2, pp. 29–36. [Google Scholar]
- Temple, S. Spinn-3 Development Board. Available online: http://spinnakermanchester.github.io/docs/spinn-app-1.pdf (accessed on 1 July 2021).
- Temple, S. Spinnaker Links. Available online: http://spinnakermanchester.github.io/docs/spinn-app-7.pdf (accessed on 1 August 2021).
- Christensen, D.V.; Dittmann, R.; Linares-Barranco, B.; Sebastian, A.; Gallo, M.L.; Redaelli, A.; Slesazeck, S.; Mikolajick, T.; Spiga, S.; Menzel, S.; et al. 2022 roadmap on neuromorphic computing and engineering. Neuromorphic Comput. Eng. 2022, 2, 022501. [Google Scholar] [CrossRef]
- Mori, H.; Satake, T.K.M.; Matsuoka, T. Communication System with a Plurality of Nodes Communicably Connected for Communication Based on nrz (Non Return to Zero) Code. U.S. Patent 20,120,051,241, 3 January 2012. [Google Scholar]
- Brouwer, A.; Shearer, J.; Sloane, N.; Smith, W. A new table of constant weight codes. IEEE Trans. Inf. Theory 1990, 36, 1334–1380. [Google Scholar] [CrossRef]
- Russo, N.; Huang, H.; Nikolic, K. Live Demonstration: Neuromorphic Robot Goalie Controlled by Spiking Neural Network. In Proceedings of the 2022 IEEE Biomedical Circuits and Systems Conference (BioCAS), Taipei, Taiwan, 13 October 2022; p. 249. [Google Scholar]
- Lichtsteiner, P.; Posch, C.; Delbruck, T. A 128 × 128 120 db 15 us latency asynchronous temporal contrast vision sensor. IEEE J. Solid State Circuits 2008, 43, 566–576. [Google Scholar] [CrossRef] [Green Version]
- Huang, X.; Li, Z.; Xiang, Y.; Ni, Y.; Chi, Y.; Li, Y.; Yang, L.; Peng, X.B.; Sreenath, K. Creating a Dynamic Quadrupedal Robotic Goalkeeper with Reinforcement Learning. arXiv 2022, arXiv:2210.04435. [Google Scholar]
- Katz, B.; Carlo, J.D.; Kim, S. Mini Cheetah: A Platform for Pushing the Limits of Dynamic Quadruped Control. In Proceedings of the 2019 International Conference on Robotics and Automation (ICRA), Montreal, QC, Canada, 20–24 May 2019; pp. 6295–6301. [Google Scholar]
- Natale, L.; Bartolozzi, C.; Nori, F.; Sandini, G.; Metta, G. ICub. arXiv 2021, arXiv:2105.02313. [Google Scholar]
- Wise, M.; Ferguson, M.; King, D.; Diehr, E.; Dymesich, D. Fetch and Freight: Standard Platforms for Service Robot Applications. In Proceedings of the Workshop on Autonomous Mobile Service Robots, New York, NY, USA, 11 July 2016. [Google Scholar]
SpiNNaker Link | 1 | 3 | 5 | 7 | 9 | 11 | 13 | 15 | 17 |
Arduino Due | GND | GND | GND | GND | GND | GND | GND | GND | GND |
SpiNNaker Link | 2 | 4 | 6 | 8 | 10 | 12 | 14 | 16 | 18 |
Arduino Due | 8 | 7 | 6 | 5 | 4 | 3 | 2 | 10 | GND |
SpiNNaker Link | 19 | 21 | 23 | 25 | 27 | 29 | 31 | 33 | |
Arduino Due | 29 | 28 | 27 | 26 | 25 | 24 | 23 | 22 | |
SpiNNaker Link | 20 | 22 | 24 | 26 | 28 | 30 | 32 | 34 | |
Arduino Due | GND | GND | GND | GND | GND | GND | GND | GND |
Arduino Due | Board | ||||
---|---|---|---|---|---|
Pins | N Pins | Type | Voltage | Shifted to | |
SpiNNaker | 2–8, 10, 22–29 | 16 | Digital | 3.3 V | 1.8 V |
Servo | 39 (data), Vin, Gnd (power) | 1 | PWM | 3.3 V | n/a |
Touch Sensor | 13 (data), Vin, Gnd (power) | 1 | Digital | 3.3 V | n/a |
DVS | USB | n/a | Serial | 5 V | n/a |
Power (batteries) | Vin, Gnd | 2 | DC | 15 V | 9 V |
Symbol | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | A | B | C | D | E | F | EOP |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
bit 1 | 0 | 1 | 2 | 3 | 0 | 1 | 2 | 3 | 0 | 1 | 2 | 3 | 0 | 1 | 2 | 1 | 5 |
bit 2 | 4 | 4 | 4 | 4 | 5 | 5 | 5 | 5 | 6 | 6 | 6 | 6 | 1 | 2 | 3 | 3 | 6 |
Name | Description | Vision Sensor | Power Consumption | Positioning Time | Accuracy |
---|---|---|---|---|---|
Quadrupedal Robotic Goalkeeper [22] | The Intel camera is used to track the target ball and send the prediction to the Mini Cheetah. A GPU is used to train the model using the YOLO algorithm. | Intel RealSense D435i | 120 W/h (Mini Cheetah max) [23] | 0.5 s/4 m field | (sidestep) ∼65% (full) ∼85% |
iCub v1.0/v2.0 (Intel ATOM D525) [24] | Humanoid robot with an embedded pc. It is composed of different actuators to simulate human motions. | PointGrey Dragonfly v2 (640 × 480 30 fps) | 288 W/h (960 W/h peak) | n/a | n/a |
Fetch (and Freight) (Intel i5, Haswell) [25] | Fetch robot is a mobile manipulator to catch and move objects (until 6 kg) | Primesense Carmine 1.09 | 20 W/h (36 W/h peak) | n/a | n/a |
spiNNaLink (this work) | Interface Board Platform to link an MCU with a SpiNNaker board. A DVS camera is used to reveal the ball direction, maintaining a low information rate and low power consumption. | Dynamic Vision Sensor 128 | ∼7 W/h (Whole system max) | 0.150 s/1 m field | 75% |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Russo, N.; Huang, H.; Donati, E.; Madsen, T.; Nikolic, K. An Interface Platform for Robotic Neuromorphic Systems. Chips 2023, 2, 20-30. https://doi.org/10.3390/chips2010002
Russo N, Huang H, Donati E, Madsen T, Nikolic K. An Interface Platform for Robotic Neuromorphic Systems. Chips. 2023; 2(1):20-30. https://doi.org/10.3390/chips2010002
Chicago/Turabian StyleRusso, Nicola, Haochun Huang, Eugenio Donati, Thomas Madsen, and Konstantin Nikolic. 2023. "An Interface Platform for Robotic Neuromorphic Systems" Chips 2, no. 1: 20-30. https://doi.org/10.3390/chips2010002