Problem

Big tech companies have invested millions into collecting map data and are creating “centralized data monopolies.” The monopolies have allowed these companies to limit who has access to essential geospatial data and how the data are used. These obstacles have created various issues preventing the map market from reaching its full potential. Some of these challenges include the following:

  • Data silos that prevent billions of dollars worth of new services - Modern digital services depend on data, making data sharing essential for innovative services. The UK government estimated that open geospatial data sharing could generate between $7 billion and $14 billion in additional revenue from new products and services in the UK market.

  • High-cost burdens on cities and businesses - Tech giants like Google are exploiting their stronghold on geospatial data, with Google Maps pricing increasing over 1000% in 2018. This price gouging has led the U.S. Justice Department to investigate the Alphabet Inc. enterprise for unfair business practices, like prohibiting the use of Google Maps with outside software and services, according to Reuters.

  • Unfair compensation for user data - Morgan Stanley estimates Google Maps will earn more than $11 billion in revenue in 2023, profits generated mainly by Google Maps’ user data. On average, analysts calculate that people lose about $500 annually by voluntarily providing personal data to tech companies. This financial loss is expected to surge to $20,000 by 2034 as data increases in diversity and value.

The problems will get worse in the future as mapping technologies gain more sovereignty over our lives. Today, maps serve us as oracles, helping us with strategy (e.g., the best route possible) and leaving us with the execution. With the emergence of autonomous cars and robots, maps become “agents” in charge of both strategy and execution. Finally, they will become “sovereign” when a centralized player like Google controls all our autonomous agents. It can navigate us through their desired routes, force us to see what they want, and make us believe this is all our free will.

To counter big tech’s data monopolies, people, businesses, and governments need a way to gather their own mapping data. Yet this is almost impossible. The high costs of data collection and technical privacy barriers prevent individuals and organizations from gathering and generating their own geospatial data.

The Challenge and Opportunity of Cameras: A 'Super Sensor' for Dynamic Maps

The camera is a commonly available sensor. There exist 1 billion closed-circuit television (CCTV) cameras or IP cameras, and an additional 44 billion cameras in the form of mobile phones, drones, cars, etc., worldwide. Combined with AI accelerated chips and computer vision AI, the cameras turn into “super sensors” capable of mining a wide range of real-time data. They can detect cars, humans, bicycles, or potholes and understand complex events such as overcrowdedness, traffic congestion, and more. Considering the drastically lower cost of cameras compared to other visual sensors (e.g., LIDAR), they are preferred by many industry leaders such as Tesla for environmental sensing.

Although attractive, using cameras to gather data from public spaces has challenges. The first challenge is personal data privacy. Citizens do not like being surveilled, and the data privacy laws — e.g., General Data Protection Regulation (EU GDPR), California Consumer Privacy Act (CCPA), and China’s Cybersecurity Law (CSL) — have become ever more stringent. The increasing stringency of data privacy regulations leads many companies to avoid launching computer vision (camera) projects because of the risks involved in non-compliance fines.

Camera infrastructure costs represent the second significant barrier to using fixed cameras as a source of data for dynamic maps. As we noted, the current tally of more than 1 billion active global CCTV cameras has cost more than $6 trillion to install, operate and maintain. Hardware or software maintenance often leads to downtime, which results in costly or detrimental data gaps.

Last updated