GIS and Digital Mapping for Better Decision Making

satellite image for GIS use

We all have heard about the cartesian coordinate system used to plot all sorts of curves and shapes. Let’s say we had to plot some features on a certain geographical location. How would we do that? Will the standard coordinate system still be feasible? What exactly, does geography look like in the digital age? The answer- GIS or Geographic Information Systems

What Is GIS?

A textbook definition of GIS or Geographic Information System would be that it is a computer-based information system designed to input huge volumes of spatial data and allow us to store, analyze, model and retrieve information as per user-defined applications.

An illustration of a map for Geographic Information system

Geographic

Refers to the different locations on the Earth’s surface

An illustration of a box containing paper docs for Geographic Information system or GIS

Information

Cartographic, photographic, digital and attribution data

An illustration of computers and phones for the system component of GIS

System

A processing system from supercomputers to mobiles

Modern industries produce enormous amounts of data. As most of these datasets contain data points about events and points of interest that are happening or are situated at some location, at many levels this data is directly or eventually tied with geography. This breed of data is on the rise ever since our data collection and storage capabilities have improved. Research estimates that over 6 billion smartphone users are looking at some form of data, and over 50 billion smart devices are interconnected to collect, analyze, and share data in 2020 alone. The pioneers of big data are able to leverage this connection.  However, most of the time organizations forget or don’t know how to utilize the locational aspects of the data. As a result, a high potential information source remains untapped. This is where GIS comes into play.  

Role of GIS in today’s world

GIS provides the tools required to ingest, analyze geographical data to produce key insights. Directly or indirectly, these systems are powering a lot of applications that we use on a daily basis. Delivery services, Supply chains, Fraud detection, Traffic optimization are some of the many applications, which are possible by connecting the dots on the map with powerful GIS systems. The ultimate aim of a geographic information system is to help make decisions efficiently: It presents us with data in a form that is directly or indirectly usable for required applications. It helps us find and understand spatial patterns and relationships.

History of GIS: Spatial Analysis Begins

In the 1830s, the second cholera pandemic had spread like a wildfire. Europe was crippled. Soon a huge volume of data was generated related to the pandemic and patients. The data was warehoused in the most manual manner. It was at this point that the locational parameter of the data was used to tackle this pandemic. The data could be used through two methods one that didn’t look at the geographical aspect of the data and the other which made use of the locational mapping in the data.

No geography approach

Essentially you have data of all patients and about when they arrived at which hospital and other details. Further, there is attribute data about the hospitals, staffing, equipment and capacity. Definitely, this data is very helpful in making predictions and decisions on how to add staffing and supplies to the hospitals. But, how far can this analysis get you without a map?

GIS approach

The problem solvers started putting the same data on a map by using the patient’s home address. Charles Picquet, a cartographer, generated a map that represented the outbreak across the capital Paris. Following Picquet’s footsteps, John Snow used a similar technique to map the cholera outbreak in London. But, that was not enough; apart from generating maps blotted with colors. Then, in 1854, a breakthrough happened – Layering of data from different geo-sources. John Snow layered the patient’s home address data with streets, property boundaries and water lines. 

A pattern emerged! Open water lines areas were shaded with high patient counts. Soon, open drains were getting closed. Focus on water sanitation standards was paid attention to.

However, it wasn’t until the 1960s when Roger Tomlinson developed the Canadian Geographic Information System for the Canadian Government that GIS tools came on their own in the public domain. It was because of this contribution of Roger Tomlinson, that he is now regarded as the Father of GIS.

What kind of data GIS creates

One of the main driving forces in the development of Digital Mapping and GIS is the sheer amount of organized and unorganized data companies have at their disposal. The power to analyze and visualize big data is what differentiates the rate of progress of companies. And this power depends upon the quality and attributes contained in the data. 

So what kind of data are we talking about here? 

Well, spatial & non-spatial data are the two kinds of data used in the GIS domain. 

Spatial Data

Anything related to specifying a location comes under spatial data. We usually use the word geospatial data to denote this category.  This type of data is sourced in the geography and engineering of the areas or objects of interest. The main parameter in this kind of data might consist of coordinates, parcel addresses, geofences, points of interest, areas and volumes of a particular locational feature and other attributes that are directly concerned with the location. E.g. location of properties or events, etc.

The Spatial Data is broadly categorized into vector and raster forms of data :

Vector DataRaster Data
These record spatial information as a single or series of x & y coordinates.It is an array of uniformly sized pixels.
These don’t look pixelated when viewed closely.Raster data looks pixelated when zoomed-in.
Vector data is used to represent boundaries or edges via points, lines or polygons.Raster data format is used to describe the interiors of areas.
In general, vector data structures take up less space.Raster data usually takes up more space than Vector data.

Non-Spatial Data

Non-spatial data comprises those data points that are not directly related to geography.  However, this shouldn’t imply any absolute independence from geographical variables. These data points are usually indirectly connected to geography. When non-spatial data is synced with a layer of spatial data, the value of insights that can be thus generated is extremely high. E.g. population, land-use, consumer-behavior at retail chains, and so on. 

Use cases of GIS

To say that GIS has a variety of applications would be a wild understatement. It has a widespread and growing application in almost all spheres of human commerce and activities. From performing rescue operations to selling goods, GIS has come a long way. 

Unperceivable to the average consumer, data and GIS analysis drives retail companies. Apparel companies use it to optimize profits and growth when opening retail stores and stocking warehouses. Insurance companies utilize GIS technologies to assist in underwriting and perform better risk management. This saves time as companies can harness the power of GIS without actually assessing the property manually by visiting onsite. Insurers can analyze disaster-prone areas with the help of location intelligence and hence, charge accordingly. The landscaping industry uses GIS to remotely measure the dimensions of lawns and gardens. The police employ GIS to gather intelligence which helps in reducing crime. Logistic companies apply GIS to make the process of storing and transporting goods more efficient.

These are just a few examples of how GIS is affecting the way we conduct businesses or perform operations. And we should remember that the entire gamut of applications of GIS is massive. 

Drone image of a town in Malaysia showing the use case of GIS
Drone image of a town in Malaysia. Imagery like this can be used to extract building footprints to help insurers in underwriting and real estate agencies for valuation. Photo Credits: Yong Chuan Tan, Unsplash

Are GIS and cartography the same?

Short Answer: No.

Is respiration the same as breathing? No. Breathing is simply the process of inhalation and exhalation. On the other hand, respiration is the biochemical process involving the breakdown of glucose into energy.

In the same way, GIS is different from cartography. Cartography simply refers to the study and process of map-making. However, GIS is a far richer concept, which builds on map making and utilizes several parameters to make our tasks more efficient. Its applications are huge, varying from industry to industry.

How GIS is affecting business decisions

One can imagine how hard it had been to manually compile, plot and analyze data from different geo-tagged sources to figure out the various geographical relationships and patterns.

Fast-forwarding to 2020, we now have a plethora of geo-datasets which can be used for this. In fact, Google Maps’ API  provides a rich dataset of roads, geo-tagged shops and residences, etc. It’s easier more than ever to connect your data source with the known data sources to derive deep insights and make impactful decisions. Let’s see how a typical business makes use of GIS. In this case, we will use the example of sneaker making businesses who want to decide on opening their franchises in the most suitable area.

Say a leading Footwear company wants to open a few retail stores. Will they take the risk and establish them arbitrarily anywhere? There’s too much at stake here, right?

Think of the problem as the cliche example given to Middle School students when the concept of the median is taught. They say if you want to open up a shoe factory, you’d like to produce a higher number of shoes that is close to the median size of footwear people wear, so as to avoid losses.

Opening up retail stores in 2020 is no different, except for the fact that it’s a bit more complex due to the fact that now we have a whole lot of data to work with. Data related to past online customers’ addresses, average expenditure data of families on footwear and of course the common shoe sizes people buy can be utilized and layered on top of one another to help optimize the whole process of opening up a showroom at the most suitable location.

With the help of geomarketing application of GIS, heat maps of supply, demand and consumer behavior can be drawn which can help businesses to drive sales and growth with confidence. Image credit- Geoawesomeness

4 steps for integrating GIS with your business

Utilizing GIS for any application is not a simple process. In summary, GIS is basically a remote data collection and analysis tool. There are numerous stages that require various kinds of expertise and resources. From satellite imaging to Artificial Intelligence and database developments, GIS encompasses a lot many skills and disciplines. And it is this meshing of different areas of human expertise that give GIS the pristine power of deriving invaluable insights from geodata. This process of using GIS to help your business can be broken down into the following stages.

1. Capturing geospatial imagery

The first step involves identifying what issue we are trying to deal with. The issue could be anywhere from insurance underwriting to mapping the number of Covid-19 cases in an area. For this, we procure the appropriate geospatial imagery along with contextual non-spatial data so that these can be analyzed to reach a viable solution. Geospatial Imagery comprises a variety of images of the Earth’s surface and the objects on it using various vehicles like Satellite, drones, and airplanes. Satellite Photographs such as those captured by NASA and Airbus,  aerial imagery- obtained via airplanes and other airborne vehicles, and drone imagery captured using terrestrial drones- are very rich sources of information of the ground-truth in our area of interest. 

2. Understanding data needs and imagery parameters

Geospatial imagery has many attributes and characteristics which need to be understood before one can start using them. For instance, one of the important imagery parameters is Image Resolution which gives us a measure of how distinguishable tiny objects are from each other. Ground Spatial Distance is the value that gives us an idea about image resolution – an image with a resolution of GSD = 1 m means that we cannot distinctly identify artifacts with dimensions less than 1m x 1m. The resolution needs to be chosen depending on the features we need to extract. There are numerous such characteristics of imagery that one needs to understand before selecting the right one for a project.  To know more about these imagery characteristics and how to decide on an imagery, you can read this article

3. Processing imagery through feature extraction

Once suitable geospatial imagery is obtained for the project, various kinds of processing techniques are applied to derive usable information from it. It basically means vectorizing the raster geospatial imagery to obtain a dataset that can be analyzed, further processed and ingested into professional software or calculations.

Identical vectorized elements are part of a layer. Numerous such layers containing different information parameters are extracted and this step is commonly known as feature extraction.  Sometimes, datasets like DSM, DEM are collated with the imagery to extract attributes that might not be possible from simple imagery. The next step after extraction is optimizing the accuracy of the data through a quality check.

Image processing is done via numerous techniques and technologies. Softwares like ArcGIS from ESRI, QGIS, etc. are traditionally very popular at drafting and extracting information from geospatial imagery. However, with the advent of AI and techniques like machine learning and computer vision, feature extraction has been highly sped up and automated. 

4. Deploying output data

Once the image processing is done, the resulting data can be exported in a wide variety of industry-standard formats known as GIS file formats. A GIS file format is the standard method of encrypting geographical information derived from various geospatial sources into a computer-readable file.  These file formats can be both raster and vector types as enumerated below.

Raster file formats

These formats represent digital images through the use of grids. The raster formats typically contain a series of rows and columns of cells wherein each cell contains definite information. These formats are usually created by governmental mapping agencies or by GIS software developers. The most popular raster GIS file formats are collected in the table below.

NameExtensionDescription
ARC Digitized Raster Graphics or ADRG.dlg or .optA popular raster file format developed by the National Geospatial-Intelligence Agency
American
Standard Code for
Information Interchange
.ascASCII uses a set of numbers between 0 and 255 for storing GIS information
Enhanced Compressed ARC Raster GraphicsAs per the ArcGIS productECRG uses the ARC system that divides the globe into 18 latitudinal zones.
Geo Tagged Image File Formats.tiffGeoTIFF is a TIFF variant enriched with GIS relevant metadata
ERDAS Imagine.img, .ige, .igwContains digital scan of a paper topographic map

Vector file formats

Vector file formats use geometrical shapes to represent different features found in an imagery. The basic elements of vectorization are as follows-

i. Points

Points are non-dimensional representations that can be used to represent those features that can be referred to by a single-point reference. E.g. POI’s, peaks, GPS locations, etc. 

ii. Lines

Lines are one-dimensional objects that are used to represent those features which have a linear geometry on the imagery. E.g. fences, boundaries, rivers, etc.

iii. Polygons

Polygons are two-dimensional objects that are used to represent, enclose, or demarcate areas on geospatial imagery. E.g. Lawns, shrub beds, footprint area of a building, solar panels coverage, etc.

The most popular vector file formats are depicted in the table below.

Name ExtensionDescription
Digital line graph.dlg, .opta USGS format for vector data
GeoJSON.json, .geojsona lightweight format based on JSON
Keyhole Markup Language.kml, .kmzXML based open standard
Measure Map Pro .mmpXML data format to store GIS data
Shapefiles.shp, .shx, .dbfa popular vector data GIS format, developed by ESRI

Is your business GIS-smart?

GIS is being used by many business and governmental organizations to improve their value chains from sales to operations to customer success. It is inevitable that using GIS will provide a better competitive edge as well as efficiency. In this scenario, the questions you need to ask about your business are:

  • How much of my data is connected with geography?
  • How can I connect my data with more geo pointers?
  • What new insights can I get, if this geo-connected data can be analyzed?
  • What are the other layers of geodata that can be added to power up these insights?
  • Do I need to invest in a GIS solution?
  • How will GIS influence my day-to-day business deals?

These questions will help you to align the benefits of GIS with your business objectives. However, this is just the first step. Implementing a successful GIS system needs lots of skills and experience. And to simplify this adoption process, we invite you to get in touch with us for a free consultation. 

At Attentive AI, we understand your business needs and with the help of our GIS and computer vision experiences, we help you to grow in all directions spatially and otherwise. Contact us for more information. 

Tags: , ,

5 Responses

  1. Ananya says:

    Good read, precise and to the point!

  2. […] use cases. Property insights can be availed in the form of pdf reports, excel files, XML and other geospatial formats like shapefiles, GeoJSON, etc. Developer teams working on existing applications can also integrate […]

  3. […] can auto-update precise geofences, resolving this problem to a large extent. As a matter of fact, at Attentive AI, we have developed models that can extract geofences with a positional accuracy of more than 99%. Moreover, the frequency of data updates can be […]

Leave a Reply

Your email address will not be published. Required fields are marked *