What does the invasion of AI in cartography mean for professional mappers? Will it prove to be a boon or a bane?
Artificial Intelligence is practically disrupting every possible progressive industry in the economy today. It has become a common buzzword echoing almost in all places. Its infinite potential is not just a center of amazement for people but also a point of intimidation for many. You’ve probably heard or read it somewhere by now about the apparent threat that looms on human jobs because of the advancements happing in AI, especially if you’re a working individual. So, before we get to the soul of this blog, let’s ask the more important questions first. Is AI really an existential crisis for humans? Will it eliminate the need for human employment? If so, how?
From tech maestros to academic philosophers, AI is a topic of discussion for all. There are a few ethical concerns regarding AI that worry even the biggest advocates and beneficiaries of modern technologies like artificial intelligence and machine learning. So when it comes to answering monumental questions like the ones asked above, the answer can’t not be paradoxical. In simple words, AI is both a threat and a non-threat to humanity. It is a threat to people who are reluctant in knowing the exact functioning of AI-powered applications and is not only a non-threat but also a real advantage for people willing to learn about AI.
How Does AI work in Cartography?
Relating the same to the field of cartography, AI can be an excellent assistant to cartographers all around the world making their work 10x more efficient as compared to their usual speed. For years, cartographers have been doing the painstaking manual task of meticulously adding data from the ground and reviewing public satellite images by hand and annotating important features to map unmapped roads around the world.
Recently a team of Facebook mappers has cataloged all the missing roads from Thailand and more than 90% of roads in Indonesia with the assistance of Map With AI. Map With AI is a service deployed by AI researchers and scientists working at Facebook.
This AI-assisted service has enabled these mappers to cover thousands of miles of roads with a dramatic speed.If done traditionally, this coverage would’ve taken them an additional 3-4 years of time. The conclusion remains, AI is a boon for people willing to understand and leverage its potential.
Using AI to Map Faster and Better
The one thing that makes modern technology so amusing is its ability to learn from patterns. Map With AI also works on the same foundation. It uses a subfield of artificial intelligence called computer vision to spot complex patterns in images so that they can analyze the same type of imagery that OpenStreetMap project volunteers have worked with for years. These systems have been trained to recognize possible roads and highlight them in the mapping tool. The OSM volunteers then use their expertise to review and confirm or modify the suggestions given by AI.
This technology, however, hasn’t been deployed at a large scale except for some proof of concept. Roads in different regions vary greatly in width and color and some inconsistencies are still to be figured out. To resolve these challenges, scientists are working on better models that will capture images with enhanced resolution. Even with such advanced technology, there are some false positives and human insight, creativity, and judgment are here to make the final call.
How Deep Neural Network Meets Human Expertise
After converting the final output of the model into a format that can then be used by mappers, the AI system’s predictions serve as the basis for the mapping process. AI makes the work of the cartographers easy as it does most of the work, the cartographers just have to double-check for accuracy, fill in any gaps, and select the appropriate road type. With the help of an enhanced version of the popular OSM editing tool called iD, mappers can do their job better. It gives mappers the final say in what ends ups being on the map and also draws enough attention to under-mapped locations.
The Roadmap Ahead
With the influx of AI in ML in cartography, the road ahead looks bright for mappers. Open-source systems and tools like Map With AI and RapiD are enabling the creation of important geospatial data with more efficiency and ease. It is about time that we stop being afraid of the unknown and ethically explore the territories of AI for a more advanced world. It is an inspiring time to be a cartographer!
Visit AITechPark for cutting-edge Tech Trends around AI, ML, Cybersecurity, along with AITech News, and timely updates from industry professionals!