ESIP Summer 2018

19 July 2018 | Tucson, AZ

Creating a Minimum Information Framework for Drone Data

and 110+ member organizations

ESIP is supported by

www.esipfed.org

@ESIPfed | #ESIPfed

Drone Cluster

Working to make scientific data collected with drones FAIR and science drones easier to use.

Tags:

Drones, FAIR Data, Semantic Tech, IoT

Want to learn more? esip-drone@lists.esipfed.org

NEW AND NOTEWORTHY

  • ESIP Laboratories grant outcomes: Minimal Information Framework for Science Drone Data Workflows
    • OSF: https://osf.io/n6t9b/
    • Take our survey! contribute to the FAIRness of drone data: http://bit.ly/droneDataSurvey

  • Drone Sensor Data Collections (Sphex and Shongololo) Snaps (Single board computer Ubuntu applications)
    • https://gitlab.com/r4space/VTAgMonitoring

  • International Data Week November 2018
    • Science Drone Flight Week
    • Drone datathon
    • SciDataCon
    • https://rpasdm.github.io/

+ 110 member organizations

ESIP is supported by

www.esipfed.org

#ESIPfed

MAKE A COPY FOR YOUR GROUP. Remember, these slides will be rotating on monitors, so keeping the text big and draw people in with visuals. Contact lab@esipfed.org with any questions!

Outline

  • sUAS scientific data challenges
    • sUAS data is unique
    • 10 challenges
  • ESIP Minimal Information Framework project
    • Case Studies
    • Ontologies
    • MIF

sUAS Scientific Data Challenges

  • What standard sensor calibration and use procedures need to be defined and articulated?

  • What best practices regarding data post processing and error analysis methodology need to be outlined?

  • What is the minimum information that needs to be collected about a scientific sUAS data capture flight?

  • Which formats should be used to store (meta)data in?

  • Which ontologies should be applied -- or need to be developed -- for sUAS (meta)data? (what we began addressing at the VOCamp)

sUAS Scientific Data Challenges

Why Care?

Why Now?

sUAS Scientific Data Challenges

Why Care?

  • Good Science! Understand & reduce uncertainty, important for science outcomes
  • Sharing! Reproducible and reusable
  • Quicker, Better Science! Increased learning and more rapid “best practices” science development

sUAS Scientific Data Challenges

Why Care?

  • Good Science! Understand & reduce uncertainty, important for science outcomes
  • Sharing! Reproducible and reusable
  • Quicker, Better Science! Increased learning and more rapid “best practices” science development

Why Now?

  • Urgent! sUAS are an increasingly used sensor platform for the sciences
  • Momentum and support! Open science and FAIR data practices
  • Possible! Maturing of the technologies to enable and implement practices

ESIP Minimal Information Framework project

A first step towards achieving FAIRness is to both augment them with machine-readable, semantically-rich metadata, and to annotate them in ways that make their provenance (the record of the processes that created the data) explicit.

Minimum information framework: a list* of data and metadata attributes necessary for sharing and reuse

Project goals:

  • Define a high-level minimum information framework (MIF) for drone data based on case studies

  • Use MIF as backbone/testbed for preliminary drone data ontology

Approach

  • Work with drone data collectors to document their:
  • Workflows
  • Data products
  • Data needs

2) Create a Minimum Information Framework (a high level information model) of key data classes necessary for reuse

3) Refine via community feedback

4) Use as basis of ontologies, data standard.

Collecting and analyzing scientific RPAS workflows

Collecting and analyzing scientific RPAS workflows

Collecting and analyzing scientific RPAS workflows

VOCamp: https://github.com/Vocamp/dronedata

  • Ontologies to build on
    • Geolink (ontology design pattern) http://daselab.cs.wright.edu/pub/2015-geolink-ontology.pdf
    • W3C SOSA
    • Various IEEE UAV/Robot ontologies

  • Format candidates
    • Onboard, web accessible: CoverageJSON
    • Archive: NetCDF

Results: Ontology Design patterns

https://github.com/Vocamp/dronedata/tree/master/concept_maps

Results: Minimum Information Framework

A high level model of key information classes and parameters; and the relationships between those information classes

Minimum information framework (so far)

*not a list

Wyngaard, J., Barbieri, L. K., Vardeman II, C., Leahy, B., Swanz, S., Thomer, A.K. (2018). Minimal Information Framework for Scientific Data Collection from Remotely Piloted Aircraft Systems (RPAS). Poster presented at 11th plenary of the Research Data Alliance. Berlin. doi:10.6084/m9.figshare.6145739

On-going: Drone Data Survey

Do you use drones in your research or teaching? We need your feedback!

http://bit.ly/droneDataSurvey

Enviro-sensing <-> Drone Cluster

  • Ontology design pattern(s) needs more community driven work (need funding)
    • By Scientific or System Domain?
  • Community discussion around:
    • Best sampling practices
    • Best standard sensor calibration processes
    • What is the minimum information necessary to be FAIR
    • FAIR archives for “small” data/time series
    • Formats… (meta)data...
Copy of ESIP Lab 2018 - Thomer Wyngaard - Google Slides