DeepNNCar: A Testbed for Deploying and Testing Middleware Frameworks for Autonomous Robots

Matthew P. Burruss, Shreyas Ramakrishna, Gabor Karsai, Abhishek Dubey

IEEE 22nd International Symposium on Real-Time Distributed Computing, ISORC 2019, Valencia, Spain, May 7-9, 2019 2019

PDF DOI

Why This Matters

Autonomous robot systems using learning-enabled components require testing frameworks that can evaluate safety guarantees and performance tradeoffs in realistic scenarios. DeepNNCar is innovative because it provides a physical testbed platform with integrated middleware framework support, enabling systematic evaluation of different control strategies and safety architectures. The platform demonstrates practical implementation of formal safety assurance with learning-enabled systems.

What We Did

This paper demonstrates DeepNNCar, a testbed platform for deploying and testing middleware frameworks for autonomous robots using learning-enabled components. The platform integrates CNN-based steering angle prediction with safety supervisors using Simplex Architecture and weighted simplex strategies for adaptive controller switching. The system demonstrates real-time performance monitoring and resource management capabilities.

Key Results

The platform successfully demonstrated weighted simplex strategy capabilities by adaptively switching between controllers based on current speed and safety performance. The system showed ability to maintain safe operation while optimizing for performance through dynamic strategy selection. The testbed demonstrated effectiveness of middleware frameworks in managing multiple controllers and ensuring safe system operation.

Full Abstract

Cite This Paper

@inproceedings{Burruss2019,
  author = {Burruss, Matthew P. and Ramakrishna, Shreyas and Karsai, Gabor and Dubey, Abhishek},
  booktitle = {IEEE} 22nd International Symposium on Real-Time Distributed Computing, {ISORC} 2019, Valencia, Spain, May 7-9, 2019},
  title = {DeepNNCar: {A} Testbed for Deploying and Testing Middleware Frameworks for Autonomous Robots},
  year = {2019},
  pages = {87--88},
  abstract = {This demo showcases the features of an adaptive middleware framework for resource constrained autonomous robots like DeepNNCar (Figure 1). These robots use Learning Enabled Components (LECs), trained with deep learning models to perform control actions. However, these LECs do not provide any safety guarantees and testing them is challenging. To overcome these challenges, we have developed an adaptive middleware framework that (1) augments the LEC with safety controllers that can use different weighted simplex strategies to improve the systems safety guarantees, and (2) includes a resource manager to monitor the resource parameters (temperature, CPU Utilization), and offload tasks at runtime. Using DeepNNCar we will demonstrate the framework and its capability to adaptively switch between the controllers and strategies based on its safety and speed performance.},
  bibsource = {dblp computer science bibliography, https://dblp.org},
  biburl = {https://dblp.org/rec/bib/conf/isorc/BurrussRKD19},
  category = {poster},
  contribution = {lead},
  doi = {10.1109/ISORC.2019.00025},
  file = {:Burruss2019-DeepNNCar_Testbed_for_Deploying_and_Testing_Middleware_Frameworks_for_Autonomous_Robots.pdf:PDF},
  keywords = {autonomous vehicles, middleware frameworks, learning-enabled components, safety assurance, simplex architecture, testbed platform},
  project = {cps-autonomy},
  tag = {ai4cps},
  timestamp = {Wed, 16 Oct 2019 14:14:53 +0200},
  url = {https://doi.org/10.1109/ISORC.2019.00025}
}
Quick Info
Year 2019
DOI 10.1109/ISORC.2019.00025
Keywords
autonomous vehicles middleware frameworks learning-enabled components safety assurance simplex architecture testbed platform
Research Areas
CPS ML for CPS middleware
Search Tags

DeepNNCar, Testbed, Deploying, Testing, Middleware, Frameworks, Autonomous, Robots, autonomous vehicles, middleware frameworks, learning-enabled components, safety assurance, simplex architecture, testbed platform, CPS, ML for CPS, middleware, 2019, Burruss, Ramakrishna, Karsai, Dubey