open source and free tools or library to manipulate these formats and services. Drone
data doesn’t have such sound standards for format and services. But drone manufacturers and third-party software enterprises have developed a whole set of systems
with their own formats and interfaces. As long as the user community accepts them,
they will become de facto standards with equivalent effect. For in situ sensors, the
data contains massive sequences, signals, and numbers which are often sorted into
the simple CSV (comma-separated values) format. The users have their habits in
organizing them. The standardization level is low in this domain. Refer to USGS
water database
12 ; the data are arranged by columns. Each column is a variable and
each row is an observation. The users can select the column to generate their own
CSVs (Goodall et al. 2008; Wolock 2003). Similar situation happens in manual
reports which are native tables. One requirement for operation of this strategy is the
data providers and consumers both clearly understand the definition of columns and
fields. The proof of practice shows that this scheme works very well. But along with
the development of big data and cloud computing, such scheme is a little
old-fashioned and inefficient. The collaboration between data sources and consumers are turning more closely to introduce the latest techniques and hardware to
further accelerate the speed of sourcing and reducing the barriers for new consumers
to entry.
4.3.2 Cloud Sourcing
The popularity of cloud computing and elastic storage brings profound impacts upon
all the walks of life. The high-performance computing, great availability, pay-peruse profit model, reasonable cost, reliability, stability, worry-free about
CPU/memory/disk failure, maintenance free, easy-to-use API, and intuitive user
interface are the basic reasons why so many organizations have transferred their
entire businesses onto cloud. The major commercial cloud platforms include Amazon Web Service (AWS) (Varia and Mathew 2014; Wang and Ng 2010), Google
Cloud Platform (GCP) (Cusumano 2010), Microsoft Azure, IBM Cloud, etc. (Qian
et al. 2009). These cloud platforms offer a variety of services from IaaS (Infrastructure as a Service) (Bhardwaj et al. 2010), PaaS (Platform as a Service) (Pahl 2015) to
SaaS (Software as a Service) (Dillon et al. 2010). IaaS can allocate an on-demand
virtual machine to users. PaaS gives a framework that the users can build upon to
develop or customize applications. SaaS delivers the applications managed by thirdparty vendor and accessible on the clients’ side. Most SaaS applications can run
directly from a web browser, requiring no installation or downloads. SaaS is the most
prosperous business in the cloud market (Cusumano 2010). Storage service is an
important piece in the cloud puzzle. Amazon EC2 (elastic compute cloud), EBS
(elastic block store), EFS (elastic file system), Glacier, and S3 (Simple Storage
12 https://waterdata.usgs.gov/nwis
58
Z. Sun et al.
data doesn’t have such sound standards for format and services. But drone manufacturers and third-party software enterprises have developed a whole set of systems
with their own formats and interfaces. As long as the user community accepts them,
they will become de facto standards with equivalent effect. For in situ sensors, the
data contains massive sequences, signals, and numbers which are often sorted into
the simple CSV (comma-separated values) format. The users have their habits in
organizing them. The standardization level is low in this domain. Refer to USGS
water database
12 ; the data are arranged by columns. Each column is a variable and
each row is an observation. The users can select the column to generate their own
CSVs (Goodall et al. 2008; Wolock 2003). Similar situation happens in manual
reports which are native tables. One requirement for operation of this strategy is the
data providers and consumers both clearly understand the definition of columns and
fields. The proof of practice shows that this scheme works very well. But along with
the development of big data and cloud computing, such scheme is a little
old-fashioned and inefficient. The collaboration between data sources and consumers are turning more closely to introduce the latest techniques and hardware to
further accelerate the speed of sourcing and reducing the barriers for new consumers
to entry.
4.3.2 Cloud Sourcing
The popularity of cloud computing and elastic storage brings profound impacts upon
all the walks of life. The high-performance computing, great availability, pay-peruse profit model, reasonable cost, reliability, stability, worry-free about
CPU/memory/disk failure, maintenance free, easy-to-use API, and intuitive user
interface are the basic reasons why so many organizations have transferred their
entire businesses onto cloud. The major commercial cloud platforms include Amazon Web Service (AWS) (Varia and Mathew 2014; Wang and Ng 2010), Google
Cloud Platform (GCP) (Cusumano 2010), Microsoft Azure, IBM Cloud, etc. (Qian
et al. 2009). These cloud platforms offer a variety of services from IaaS (Infrastructure as a Service) (Bhardwaj et al. 2010), PaaS (Platform as a Service) (Pahl 2015) to
SaaS (Software as a Service) (Dillon et al. 2010). IaaS can allocate an on-demand
virtual machine to users. PaaS gives a framework that the users can build upon to
develop or customize applications. SaaS delivers the applications managed by thirdparty vendor and accessible on the clients’ side. Most SaaS applications can run
directly from a web browser, requiring no installation or downloads. SaaS is the most
prosperous business in the cloud market (Cusumano 2010). Storage service is an
important piece in the cloud puzzle. Amazon EC2 (elastic compute cloud), EBS
(elastic block store), EFS (elastic file system), Glacier, and S3 (Simple Storage
12 https://waterdata.usgs.gov/nwis
58
Z. Sun et al.