Transcript presentation - Center for Systems and Software Engineering

Tatiana Kichkaylo
with Dave Barnhart and Lucy Hoag
USC Information Sciences Institute

Funded as ISI Inter-divisional seedling
Space Engineering Research Center
+ Artificial Intelligence
= Spacecraft
Portal for
Systems Platform
forIntegrated
IntegratedDesign
Designin
inReal
Realtime
time

Driven by user needs
 SERC is building satellites, fast, cheap, with student power ("teaching
hospital" model)
 SPIDR is there to help
Spacecraft missions are typically “one-off” designs, requiring high degree of nonrecurring engineering between even similar design states
 Spacecraft design is highly interdisciplinary, and follows many sets of constraints
and rules dictated by environment, payload and mission sets

Components and design approaches can be reused
Engineering knowledge can be encoded to automate many design choices and
analysis steps
 "Paperless" design helps directly and also provides basis for further automation


Magic

Blueprints
Blueprints
Blueprints
Purchase
Purchase
Purchase
orders
orders
orders
Launch dates
Launch dates
Launch
… dates
…
…
Ultimately want to let individual engineers bounce ideas of a computer system,
leading to faster, cheaper and better designs
Flight software
Mission goals
Geometry/mass
Delivery schedules
Design
Power
Launch
Component providers
Payload
Orbit
Manufacturing capabilities
Vibration
Thermal
Validation and testing
(software and hardware)
Export regulations

Collaborative design
 Information sharing and clash detection

Optimization of parameters
 Limited number of parameters but with high fidelity

Project management
 Schedules and resources

SPIDR
 Automate roll out and propagation of dependencies
 Optimize over structures given high-level specification
 Very lightweight and driven by project needs

AI planning is about finding a sequence (DAG) of actions that
achieve a goal
 Methods have preconditions
 Methods expand parts of the plan

Design is about finding a set of recipes that convert a
specification into a complete design
 Recipes have applicability conditions
 Recipes modify (expand) design


Both methods and recipes have parameters
Both plans and designs have constraints
 Constraints in design are often represented as models
Reusable data
Reusable
component
database
Reusable
knowledge base of
engineering rules
Mission-specific data
Mission
requirements,
orbit
Optimization
metrics
SPIDR
Automated design
synthesis
N best designs
(component lists and
system budgets)
User revises requirements
and metrics as needed
Optimization
engine
SPIDR
User interface
Planning layer
Planner
Constraint layer
Variables and
constraints
Constraint
propagation
Semantic layer
Semantic
Query engine
entities
Data
representation Algorithms
iLands
Tokens and
rules
Constants
P. battery
Mission req's

Nanosatellite project at USC ISI
 Low costs
 Unusual constraints

Reality alters trajectory
 Initial "vision" for SPIDR
 Single optimization tool with models in the loop
 Actual use




Run every several days with design files (XML etc.) posted online
Got several extensions, model tweaked constantly
Useful for dealing with ripple effects and ensuring consistency
Users want more

No way to close all loops
 Extended to dynamic models (e.g., power profile)
 Stuck on thermal and geometry due to
 Lack of APIs
 Licensing issues
 Need for user interaction

One shot is not enough – need Living Design




Feed back from out-of-the-loop models, version control
Side-by-side designs
Project structure and scheduling roll out
Still want the optimization loop

Rule design is not easy
 Engineers understand the concept
 Tools help: Eclipse plug-in
 Crossing domain boundary is hard
 Tweaks and overrides encourage local thinking
 SPIDR Lite to study the knowledge engineering problem

Other applications:
Trade space studies and tools for acquisition
 SPIDR-Energy for undersea surveillance (DARPA)

SPIDR is about rolling out engineering recipes and optimizing
over high-level structures
 Glue for collaborative design, optimization, and project management

SPIDR is driven by real project(s)
 Some ambitions got scaled down
 New extensions and ideas emerged

Next steps
 New [engineering] domains
 New applications
 More users (students?)
Tatiana Kichkaylo
[email protected]