I’ve long been bothered by one question: What is the future of geospatial?
I know: This is a tough question to answer. I also know the $8 billion niche of today (Geospatial 1.0) will be but a small part that future. I keep reading about disruption. Challenges. Innovation. Opportunities. Leadership.
Without meaning to sound harsh, words are one thing, action quite another.
So I have been engaged in a voyage of discovery: writing articles, conducting webinars and gathering expert opinions: Connecting dots, in an effort to build out an action plan.
Geospatial 2.0 is a key part of the geospatial industries future
I have little doubt that Geospatial 2.0 will be a huge disrupter. I’ve introduced it at a high level in past articles and webinars. But at present it is a convenient label, the details require more fleshing-out.
In this article, I want to make the first in-roads in more clearly defining Geospatial 2.0. I will here provide answers to two questions:
1) What are the elements of Geospatial 2.0?
As the diagram above shows, the architecture of Geospatial 2.0 is made up of four main elements:
a) Raw Data (device/sensor)
These are the data collectors. This includes data collected by satellite, aerial, drone, and terrestrial (IoT) methods. Imagery, video and LiDAR are among the types of data being collected.
Venture capital firms are becoming increasingly more interested in investing in these types of companies.
b) Data Fusion (AI Aggregation)
Data is far more powerful when individual sources are aggregated or fused. In the online world companies like Netflix, Facebook, and Google aggregate user information (avoiding the user data silos most web analytics methods create) and use AI to provide individualized recommendations to users. This is an automated process called a System of Insight (see the Forrester report – Digital Insights are the New Currency of Business).
A similar process can be applied to location-based data. That starts with fusing data together as a pool. AI is often used to aggregate data sources. Pixel8.earth, a Boulder based company, are doing just this with 3D data.
According to co-founder Sean Gorman:
Pixel8 use computer vision and AI to covert and fuse imagery into a 3D data source.
Machines replace humans. Geospatial data has its own very unique challenges. Groups like Pixel8 are simplifying the processing and fusing of data. The data pools generated provide the raw materials or fuel for analytics.
c) AI (Analytics-as-a-Service)
AI machine learning algorithms can answer very focused questions by leveraging large datasets. Training is required to ‘teach’ an algorithm. That might mean, as we show below, teaching an algorithms to identify garbage by the side of a road.
AI algorithms are many and varied. We read much in the autonomous vehicle world about AI identification/classification: a car verses pedestrian for example. Predictive AI is a burgeoning new field. My group have been doing some work around predicting slope failure using AI and radar data. We are exploring if we can predict slopes most vulnerable to failure? One fascinating new emerging field is called prescriptive AI. That is, taking predictive results and changing variables to affect the outcome. Take for example a flight which AI predicts will be 5 minutes late, how might that arrival time be improved if the arrival gate were changed?
For this description, Analytics-as-a-Service is a convenient way to describe a basket of AI (or machine learning) services.
d) Geospatial Foundation (Framework)
We will cover this in more depth in the next section. But the Geospatial Framework is essentially the glue which pulls all the Geospatial 2.0 elements together. It is both the foundation and visualization engine.
Aerial observation by satellites and drones combined with artificial intelligence could help tackle a wide variety of challenges.
2) Is there a Geospatial 2.0 framework in place which will help drive adoption and innovation?
Discussion and collaboration is needed around the evolution of a Geospatial 2.0 framework. We have groups collecting and processing data. Others building AI algorithms. Now a framework is needed which connects each of the core Geospatial 2.0 elements.
There are many good proprietary Geospatial 1.0 frameworks in existence. Many use the word ‘open’ to describe their frameworks. They are certainly open to their subscriber base. But, empowering 2.0 innovation will require ‘open to all’ frameworks. That does not exclude the 1.0 world, but moves 2.0 out of the hands of any controlling interest, stimulating innovation without constraints.
Though not promoting one or other existing solution: Data61, a public entity in Australia, have been building what I see as potentially a good starting point for a Geospatial 2.0 initial framework. See this video: https://youtu.be/LiDW5v9RdM0
There does not currently exist a widely supported open geospatial framework. As I will discuss below, there need be cooperation between global public entities to help evolve this important foundation.
For wide Geospatial 2.0 adoption, a geospatial framework is a core requirement. This will bring disconnected groups together, stimulate innovation and spawn a new set of solutions, services and Geospatial 2.0 companies.
Private and Public Sector Synergy ..
The evolution of Geospatial 2.0 will require cooperation and a close relationship between the private and public sectors. Three reports have been released which discuss the future of geospatial:
- UN-GGIM Future Trends Report
- Geospatial Commission Strategy Report
- Knowledge Transfer Network/Ordnance Survey Power of Place Publication
Stimulating innovation forms key parts of each of these reports. On this topic, it is my belief that there are two steps urgently needed to move these reports from strategy documents to action:
1) Open Inclusive Geospatial Framework
Collaboration, discussion, agreement and funding are needed to build out an Open Inclusive Geospatial Framework. Above I mentioned the work done by Data61. They have leveraged two popular and well supported open source technologies: Leaflet and Cesium.
Extending this type of publicly supported open framework is essential.
2) International Innovation Network
Having the technical infrastructure in place to enable Geosptial 2.0 is one thing. We now also need to connect the players: To drive synergy between solution providers. Geovation in the UK provide a template. We need a network which brings complimentary companies and solutions together. Not only potentially providing office space, but a knowledge network. That is both technical and business-centric. I have been working with some public entities here in the US to move this need forward.
A wider set of players is needed to collaborate and help make this vision a reality
In Summary ..
These are exciting times. Geospatial data will soon be empowering decision makes across all sectors. But there is still work to be done to enable Geospatial 2.0.
Information is the lifeblood of financial markets. The systems that collect, collate, and disseminate financial market information are a key component of well-functioning capital markets. Geospatial data and analysis will have profound implications for these information markets and associated systems.
We have discussed the core elements of Geospatial 2.0 in this article. We have also covered the immediate needs to move Geospatial 2.0 closer to reality.
Much work is still needed.
Collaboration will be the key to moving Geospatial 2.0 forward. I hope in the near future to be writing progress reports: articles which summarise collaborative meetings and Geospatial 2.0 advances.
This article is in many ways my call to action to the geospatial community. Let’s join together and stop imagining the future. Let’s instead together create it.
You can reach me on [email protected]