Artificial intelligence (AI) and Geospatial science

"We are not thinking machines that feel; rather, we are feeling machines that think." – Antonio Damasio.

A human being is the most developed and superior creature on this planet. We have developed our intelligence and knowledge like we did never before. We are always fond of intelligent machines and this encourages us to make some super-intelligent machines. 

Artificial intelligence (AI)

This is the intelligence that is demonstrated by machines, unlike human intelligence which is developed by the human mind. The term artificial intelligence was first used in 1965  by John McCarthy. In simple words, we can define it as 'an ability of a machine or a computer program to think and learn. The basic idea is of making machines that can think, act and learn just like a human mind.'

Machine Learning

This is one of the applications of AI. It is about making systems capable of learning and improving themselves without being programmed again and again. This is possible by developing new computer programmes. Machine learning enables analysis of massive quantities of data. Such programmes are faster and more accurate in order to identify profitable opportunities or dangerous risks, it may also require additional time and resources to train such professionals.

Type of Artificial intelligence

Artificial intelligence can be categorized in the following two ways:

Weak AI 
This is also known as narrow AI. This is programmed and designed for a specific task. Apple's Siri is an example.

Strong AI
This is also known as artificial general intelligence. This is an AI system with generalized human cognitive abilities so that when presented with an unfamiliar task, it has enough intelligence to find a solution.

As per Mr. Arend Hintze, an assistant professor of integrative biology and computer science and engineering at Michigan State University, AI can be categorized in the following four ways:

1) Reactive machines
An example is Deep Blue, the IBM chess program that beat Garry Kasparov in the 1990s. This computer program is capable of identifying pieces on the chess board and can make predictions. It analyses its own and its opponents move too but it has no memory and can't use the data for future.

2) Limited memory
These are the kind of programs which can use past experience and use them in future. Some functions of autonomous vehicles are based on this. Their observations are not stored permanently.

3) Theory of mind
This is a psychology term. It refers to the understanding that others have their own beliefs, desires, and intentions that impact the decisions they make. This kind of AI does not yet exist.

4) Self-awareness
In this category, AI systems have a sense of self and consciousness. Machines with self-awareness understand their current state and can use the information to infer what others are feeling. This type of AI does not yet exist.

Examples of AI

Robotic process automation (performing the high volume repeatable task)

Machine learning (computer act without programming)

Machine vision (making the computer see using the camera and then analyzing the data)

Natural language processing (NLP), best-known example is spam detection for e-mails.

Robotics (Design and manufacturing of robots)

                                                 Robots assisting in car manufacturing

Artificial intelligence for geospatial data customers and suppliers 

Over the next decade, the demand and supply both are going to increase for geospatial data. Some of the key factors behind this fact are – Far better remote sensing data capture techniques are developing like improved sensors, UAV (Drones) and aircrafts equipped with the highest quality of sensors, open access of data is increasing and we are developing new and robust algorithms and machine learning.

The major challenge is expressing and attributing the features of data that would be more meaningful to the data consumers in the context of their businesses. It is the difficulty that both sides face while communicating regarding the data supply. Here come the machines into play. To bridge this gap and make things more simple at both supplier and consumer ends, machine learning or advanced algorithms are used, which create'vocabularies'.These vocabularies will help the supplier to define their data in a manner that enable a customer to the understand data for his application.

 

Microsoft and ESRI joined hands for Geospatial AI or Geo.AI

Geospatial AI or Geo.AI is machine learning, based on geographic components. This is basically integrating geography, location and AI. For conservation planning, we need to do the labor-intensive land cover mapping. Collecting the data with current methods is a time taking and laborious job. Technologies like AIcan help in such cases.

 

The two giants in their respective fields, Microsoft and ESRI  joined hands for empowering Geospatial AI. This will bring AI, cloud technology, and infrastructure, geospatial analytics and visualization together to help create more powerful and intelligent applications.

Push the boundaries of satellite data analysis (Geospatial object detection)

Over the past decades, considerable efforts have been made to develop various methods for the detection of different types of objects in satellite and aerial images, such as buildings, storage tanks, vehicles, and airplanes. Object detection in high-resolution satellite (HRS) images determines whether there are one or more objects belonging to the classes we are looking for and locates the position of each object using a bounding box. One of the four existing methods for object detection is machine learning-based method. For machine learning-based methods, three processing steps are needed: feature extraction, feature fusion dimension reduction, and classifier training. Objects with extremely small scales could also be detected as well, e.g., the storage tanks.
 
                                                          Object Detection results using machine learning-based method

Summary

Artificial intelligence (AI) is a modern-day technique which is making our machines smarter than before. It is not about adding more programs to the software's or reprogramming them but it is about improving the capabilities of a machine so that it can learn, analyze, memorize and can take the best decisions as per the circumstances. Machine learning is the basic concept behind it. Geospatial AI or Geo.AI is machine learning, based on geographic components. This is basically integrating geography, location and AI. This has endless possibilities.

About SATPALDA

SATPALDA is an ISO 9001:2008 certified leading provider of geospatial products and services. The company is also a reseller of multiple satellite data products and has a proven track record of delivering project-critical geospatial products, including satellite imagery, UAV imagery, elevation models, LULC maps, planimetry, terrain solutions as well as innovative solutions such as countrywide datasets and Airport Mapping solutions. With an experienced team having thousands of man-hours of project experience and expertise, the company focuses on providing value-based solutions to its customers in a time-bound manner. As a responsible social entity, SATPALDA strives to promote the latest Geospatial technologies to support critical issues such as climate change, environment, and security etc.