OCS Scheduler Status Francisco Delgado T&S Software Manager LSST Joint Technical Meeting 2017-03-07 | Glendale, CA

Automatic & Dynamic LSST Scheduler LSST as a robotic observatory Targets are Field/Filter Survey is automatic Multiple science goals Combine area distribution with temporal sampling Dynamic adaptation to weather Flexibility for survey adjustments during operations Flexibility for changes in science programs

Prototype History OpSim1 Proof of concept in IDL Simulation of visits with multiple science cases. OpSim2 Python Detailed model for the observatory and the weather Embedded Scheduler prototype Telescope design validation, site selection OpSim3 Additional science cases and scheduling algorithms Modular Scheduler prototype Parameters exploration, survey definition validation

OCS Requirements Flow Down Science Requirements Document LPM-17 Science Book Science Collaborations LSST System Requirements LSE-29 Metrics Requirements DOC-15319 Observatory System Specifications LSE-30 OpSim Requirements LSE-189 Telescope & Site Requirements LSE-60 Observatory Control System Requirements LSE-62 Scheduler Requirements LTS-347

History of Reviews 2011-08-29 NSF Preliminary Design Review 2012-05-30 NSF Joint Interface & Management Review 2013-02-06 OCS/TCS Software Review 2013-12-03 NSF Final Design Review 2014-02-04 OpSim/Scheduler Review 2014-08-01 Start of Construction 2014-09-10 OCS Interface Review 2016-02-08 NSF Status Review 2016-08-16 NSF/DOE Status Review 2016-10-20 OCS/TCS Software Status Review 2016-11-17 T&S Integration Review 2017-01-24 DOE/NSF Commissioning Review

Observatory Control System Control Context Observatory Control System TCS OCS Application OCS Scheduler CCS OCS Sequencer OCS EFD DMCS

Operations Simulator v4 Simulation Context Operations Simulator v4 OCS Scheduler Simulated OCS (SOCS)

OCS communications Middleware Control Data Flow OCS Operator OCS Remote OCS Monitor OCS Application Control Scheduler History OCS EFD OCS Sequencer Telemetry Image Quality Cmd Visits Targets Sched Telem Visits OCS communications Middleware TCS CCS DM

OCS communications Middleware Simulation Data Flow Control Scheduler History Telemetry Image Quality Targets Sched Telem Visits OCS communications Middleware SOCS

Internal Block Diagram Time Main Driver Sched Config Scheduler Sched Mode Control Targets Targets Downtime Cost functions Sched Telem Degraded Sched Telem Slew Time Observatory Model Telemetry Observatory conditions Sky Model Pre-Calc Data Candidates Weather forecast Proposals Image Quality Environment conditions Value functions History Quality parameters Observation History Past observations Visits Current observation 10

Science Proposal Classes Scripted proposals Designed to produce a pre-determined sequence Basic parameter: scripted cadence of targets Area distribution proposals Designed to obtain uniform distribution Basic parameter: goal visits per filter Look-ahead info: future available time for the targets Time distribution proposals Designed to obtain specified intervals in sequences Basic parameter: time window for visits interval Look-ahead info: visibility for next intervals

Accuracy changes Repeatability 100% repeatability in observations sequence and slew states Alt-Az precision Alt-Az estimated at slewInit and recalculated at slewFinal Difference can be seen between “target” and “slewFinalState” Remaining tracking time Taken into account before sending target (+30 seconds) Avoid reaching tracking limits in altitude, azimuth or rotator

Accuracy changes Sky brightness model New model, per band, includes explicit twilight Configurable moon avoidance New sky model also includes it (built-in) No caching ranks (reuse) Everything is ranked at each visit

Ranking a Target Hard constraints Global Filter change rate, burst and average Observatory kinematic limits Per Proposal Airmass Transparency Seeing Sky region (dynamic for rolling cadence) Per Filter sky brightness

Area distribution algorithms Need ratio

Area Distribution changes Time window hybrid Optional configurable grouped timed visits The first in the group is area ranked The followings in the group are time-window ranked Same night revisits constraint Revisits to the group during the same night can be avoided No overflow

Rank Bonuses Airmass bonus Hour Angle bonus

Ranking a Target Need Function Time Distribution Look-ahead: future sequence feasibility Value Function

Proposals priorities changes Self balancing mechanism

Slew time cost

Filter change cost

Final target rank Value Boost and Cost

Serendipity changes Proposal Id based PropId is included in the observation Winners and Losers counted If serendipity is allowed, both lists are searched Coadding is now optional Coadding values for a target in more than one proposal can be disabled

OCS Scheduler plan Scheduler v0.1 2015-July OCS Middleware framework Scheduler v0.2 2015-December Observatory kinematic model Scripted targets proposal Scheduler v0.3 2016-May Area distribution proposals Sky brightness model Airmass bonus Slew time cost function Dynamic configuration

OCS Scheduler plan Scheduler v1.0 2017-March Filter swaps around new moon Filter change rate constraints Filter change cost function Downtime handling Clouds handling Seeing handling Hour angle bonus Self balancing proposals Grouped visits for Area distribution proposals Time distribution proposals with sequences Parallel subsequences Rolling cadence Deep drilling

OCS Scheduler plan Scheduler v1.1 2017-August Deterministic look-ahead area distribution Scheduler v1.2 2017-December Deterministic look-ahead time distribution Scheduler v1.3 2018-April Speed performance improvements Scheduler v1.4 2018-October Warm start to resume partial survey

OCS Scheduler plan Scheduler v1.5 2019-March Dithering patterns Scheduler v2.0 2019-August Predicted schedule publication Scheduler v2.1 2019-December Weather forecast handling Scheduler v2.2 2020-May Community provided optimizations

(Metrics Analysis Framework) Validation OpSim (Operations Simulator) Observations Database Scheduler SOCS (Simulated OCS) MAF (Metrics Analysis Framework)

Analysis of runs

Quality Assurance Unit Test Integration Test with SOCS Algorithms Validation Validation runs by Systems Engineering Simulations Release Tests (includes installation and fiducial run) Scientific Validation by Science Collaborations Integration Tests with OCS

Software Engineering Tools Model Based Systems Engineering SysML modeling with Enterprise Architect PMCS with Primavera Coordination and progress with JIRA Documentation in Docushare Interface controlled in OCS SAL with XML Version control with Git, Stash Mirrored on GitHub Continuous Integration with Jenkins Coding standards from Telescope group and Simulations group

Summary Scheduling algorithms validated during prototype effort Scheduler designed for control and simulation context Development plan coordinated with SOCS for I&T Integration activities with Observatory Control System In Control context the Scheduler drives the LSST survey In Simulation context the Scheduler evaluates alternate surveys for parameters exploration and optimizations during operations