IoT (Internet of Things) devices are expected to continue to explode in popularity. A Gartner estimation places the number of connected devices in 2017 at 8.4 billion while growing to an estimated 20.8 billion in 2020.
Connected cars already factor into this IoT device calculation—however, the market debut of a fully autonomous self-driving car is not yet expected until the early-to mid-2020s at the earliest. Nonetheless, the impact of its entry will be significant, with analysts estimating 25% of U.S. driving being performed by self-driving cars by 2030.
IoT devices already generate billions of points of data. However, analysts predict that the average self-driving car will generate 1 gigabyte of data per second. This pushes the number of data points from billions into the trillions and beyond.
Manufacturers need to be able to access and interpret this information to enable data-driven optimisation of their vehicles. This need becomes especially pertinent when considering that, by 2050, the self-driving vehicle market will represent a $7 trillion opportunity — $3.1 trillion of which is represented by commercial delivery fleets and consumer-facing pilotless vehicle services. Corporate-level fleet owners also require the same critical insight as manufacturers to ensure their vehicles are running efficiently and safely.
Luca Garulli, OrientDB co-founder and CEO, suggests the answer is to this monolithic data problem can be found using a unique type of database: Graph Databases.
They are NoSQL databases that use the graph data model comprised of vertices and edges. A vertex is an entity such as a person, place, object, or relevant piece of data and an edge represents the relationship between two vertices.
Similar to how people naturally form links between relevant data, Graph Databases make relationships between data points a primary concern. Graph Databases are prized for their superior speed over traditional databases—on average, graph DBs are thousands of times faster depending on the query.
“The flexibility and scalability of Graph Databases is what is truly going to provide the rich insights self-driving car manufacturers are looking for,” Garulli says.
“For example, consider the array of geospatial data and geolocation data that a self-driving car would generate. Using a graph database, this data can be cross-referenced with weather, traffic and internal diagnostic datasets to understand relationships between, say, fuel efficiency and time of day. The granularity of the data means it could even be possible to measure the performance of individual tires during specific weather conditions or events. The possibilities are endless and graph databases are the best tool to help us navigate the incredibly data-rich future self-driving cars will help usher in.”
“What’s really exciting is that this information can be used to inform future design decisions. Graph Databases’ ability to identify patterns comes into play with the design aspect. Say, for example, if a common issue crops up amongst an entire fleet of cars, spotting it will be much faster and easier using a graph-based solution.”