Context Aware Software Stacks for Mobility: Composive.ai Overview
summary
A model-driven pipeline extracts a running Autoware ROS graph into node/topic models, uses rule-based and ML context detectors to score context fitness, composes alternative module configurations and deploys them with lifecycle commands (launch/load/unload). The project targets lightweight runtime on embedded platforms, plans a v1 release in three months, ships as open source under Compositive.ai, and validates the approach on a 1/10 scale RC car using a custom ROS graph DSL for adaptation.
AI-written summary · about
chapters (17)
- 0:00Intro
- 1:15Welcome
- 1:42Outline
- 2:01Background
- 3:44Autonomous Driving Approaches
- 4:56Genetic Ad Stacks
- 5:47New Approaches
- 7:36Context Variation
- 8:14Context Definition
- 9:13Context Awareness
- 9:57Modular Pipeline
- 11:02Robomosis
- 11:33Overview
- 14:13Modeling in Context
- 14:56Mobility Applications
- 15:46Roadmap
- 16:33Questions
related talks
46mFullstack AI: The Building Blocks for Production Ready AI-Powered Software
06mThe Building Blocks of Agentic Al
52mSerg Masís | AI Track | Compass AI & Data Summit 2025
21mML Infrastructure for Autonomous Vehicles
36mApplied AI on the edge
19mBuilding Data-Driven, Intelligent Apps in the Age of GenAI
38mStreamlining AI Solutions Development: S...
49mAI Composite in the Mayflower Autonomous navigation solution