Difference between revisions of "Assurance for Learning Enabled Systems"

From MurrayWiki
Jump to: navigation, search
(Created page with "{{subst:project boilerplate}} {{Project |Title=Assurance for Learning Enabled Systems |Agency=DARPA |Start date=1 July 2019 |End date=30 Jun 2021 |Support summary=1 postdoc, 0...")
 
 
Line 1: Line 1:
 
{{righttoc}}  
 
{{righttoc}}  
Project description (typically about a paragraph)
+
This project will extend previous work in automatic synthesis methods for planning and model-predictive control to address
 +
(1) real-time synthesis through efficient and incremental constraint-solving, and (2) risk-awareness by explicitly modeling uncertainty in perception and dynamics modeling. ALES includes a risk-aware planning and real-time synthesis engine to generate plans, protocols and control that are correct-by-construction. We plan to use a special class of stochastic predicates called chance constraints to express confidence in the learned component or its individual outputs.  We specify the semantics of the temporal evolution of this logical model based on an underlying learning algorithm.
  
 
{| cellpadding=0 cellspacing=0 width=80%
 
{| cellpadding=0 cellspacing=0 width=80%
Line 11: Line 12:
 
|
 
|
 
Collaborators:
 
Collaborators:
 
+
* Susmit Jha (SRI)
 
Past participants:
 
Past participants:
 
{{project past participants}}
 
{{project past participants}}
Line 17: Line 18:
  
 
=== Objectives ===
 
=== Objectives ===
[[Image:project-name.png|right|400px]]
+
[[Image:assured-autonomy.png|right|400px]]
Description of the main objectives of the project
+
SRI and Caltech shall develop algorithms for correct-by-construction synthesis from high- level contracts, and planning in presence of uncertainty.  Caltech's primary objectives are to support the following milestones:
 +
* Year 1: Develop algorithms for correct-by-construction synthesis with probabilistic notion of safety
 +
* Year 2: Extend the approaches  to incorporate risk measures such as Conditional Value-at-Risk.
  
 
=== References ===
 
=== References ===

Latest revision as of 04:19, 10 June 2019

This project will extend previous work in automatic synthesis methods for planning and model-predictive control to address (1) real-time synthesis through efficient and incremental constraint-solving, and (2) risk-awareness by explicitly modeling uncertainty in perception and dynamics modeling. ALES includes a risk-aware planning and real-time synthesis engine to generate plans, protocols and control that are correct-by-construction. We plan to use a special class of stochastic predicates called chance constraints to express confidence in the learned component or its individual outputs. We specify the semantics of the temporal evolution of this logical model based on an underlying learning algorithm.

Current participants:

Additional participants:

Collaborators:

  • Susmit Jha (SRI)

Past participants:

Objectives

Assured-autonomy.png

SRI and Caltech shall develop algorithms for correct-by-construction synthesis from high- level contracts, and planning in presence of uncertainty. Caltech's primary objectives are to support the following milestones:

  • Year 1: Develop algorithms for correct-by-construction synthesis with probabilistic notion of safety
  • Year 2: Extend the approaches to incorporate risk measures such as Conditional Value-at-Risk.

References

None to date


  • Agency: DARPA
  • Grant number:
  • Start date: 1 July 2019
  • End date: 30 Jun 2021
  • Support: 1 postdoc, 0.5 graduate student
  • Reporting: