Real-time GIS is not a new concept, but its growing potential has significantly evolved in the last 10 years. Certain technological hurdles had to be overcome before it was feasible to access and utilize real-time data, the biggest being widespread implementation of GNSS. Other key factors that encouraged the development of real-time GIS include decreased costs of sensors, increased use of cell phones, widespread Wi-Fi technology and assisted GNSS, which triangulates using satellites and pinpoints locations between provider cell towers. This infrastructure allows data to be collected and transmitted by millions of sensors to platforms that compile databases. Software is then used to analyze, compare and produce intelligent information that supports decisions, and it happens in real time.
Who Benefits from Real-time GIS?
Providing real-time driving directions is one of many applications for real-time GIS. Individuals use on-the-fly mapping capabilities every day. A reliable network street dataset with constant updates is necessary to accurately identify a driver’s location, indicate an accident ahead, and find the cheapest gas station along the route. For those who remember printing hard-copy maps from MapQuest before a trip, only to get lost because the source map was out of date, the convenience of accurate in-car navigation is much appreciated.
Looking at the big picture, real-time GIS is useful for tracking the location and status of any kind of mobile asset, for example a truck, a drone, a ship or a smartphone. Today, the primary use is gathering information about crews and commercial fleets. Autonomous vehicles also rely heavily on real-time GIS. Without a driver to respond to changing conditions, such as drifting out of a lane or icy roads, it is even more important to have accurate, up-to-the-minute information. There are many companies working together to compile very detailed transportation network data, including those using crowdsourcing, such as Waze and HERE.
Drone manufacturers are incorporating real-time GIS to allow the systems to fly at a certain altitude and track a target on the ground. Sensors also keep the drone aware of its surroundings, help it avoid collisions and prevent it from flying into restricted areas.
In addition to tracking mobile assets, there is increasing interest in data collected by stationary sensors that record a constant data flow. Some industries are also starting to combine moving and stationary sensors, like in a car. In addition to a car’s location, information such as the last oil change or low tire pressure could generate maintenance alerts for the owner. Stationary sensors are very useful for monitoring conditions in remote, dangerous or difficult-to-access places such as manufacturing environments, military and public safety situations, cell towers, pipelines, natural disasters, etc.
Analysis Software for Real-time Intelligence
Large amounts of real-time data are being collected continuously; however, the value lies in the analysis and decision support that takes place afterward. One solution is Esri’s ArcGIS GeoEvent Server. Its core job is to integrate real-time data from many sources, and compile and convert the input to an “event” of observational data. GeoEvent Server complements the Internet of Things (IoT) platforms, such as Amazon’s AWS IoT and Microsoft Azure IoT Hub, by accepting a simple feed of data, adding the appropriate geospatial information and analyzing for patterns.
The process starts with devices in the field, sensors and actuators that are built to an IoT platform standard. Those devices transmit data through a secure infrastructure provided by the IoT platform. Large IoT platforms have broad horizontal range, and can connect to many different devices and inputs. Other platforms work only with devices in specific vertical markets like water quality and agriculture (farm equipment). Esri then takes the data, runs the analysis in real time, and sends the results out to a bigger database for response and storage.
“As of June 2017, 1,094 organizations in 80 countries were using GeoEvent Server,” says Josh Joyner, product manager for ArcGIS GeoEvent Server. “The applications are numerous and varied, including assisting public safety at the Boston Marathon, monitoring activity of snow plows and sanitation trucks, and tracking weather event emergency responders.”
Data may trigger a real-time response or be stored to look for trends over time. “If we see that the speed of a vehicle is exceeding a maximum speed threshold for more than a set time, there could be a trigger to slow the car down,” Joyner says. “Another example is automating an ‘unsafe condition’ warning when temperature sensors in a warehouse hit a threshold. Working with an IoT platform, we can send a message directly to the air conditioning in that facility. Our goal is to produce actionable data that results in an actual response.”
Too Much Data?
Big data generated by millions of sensors worldwide creates a unique set of challenges. The potential exists to collect an overwhelming amount of data, so some thought is required to achieve intelligent processes. We have the capability to collect huge volumes of data, but how much do we really need? What update frequency is useful? Is there added value in collecting data every five minutes, once an hour or once a day? How long should the data be stored? At what granularity or scale is the data most beneficial?
The answers to these questions vary by application. Some want to perform comparisons over time, others are interested in a snapshot at a specific time, and others require a continuous stream of data for faster insights and better situational awareness. By considering the desired results, the ideal data frequency, scale and storage time can be determined for all kinds of operational analysis. Government regulations, internal company policies and privacy issues also have an impact on data handling practices.
To put data volume in perspective, consider that if 10,000 sensors update every second, it will result in 1 billion new records being created every 28 hours. At that rate, it would only take a little more than two days to exceed the storage capacity of traditional databases. Large data storage solutions, such as Esri’s ArcGIS Data Store, and specifically its spatiotemporal big data store, provide the capability to scale the data source to perform rapid analysis.
There is no question that the trend for more connectivity with many types of devices will continue. We will see more smart towns/communities and other organizations using real-time GIS to achieve operational cost savings and quality of life improvements. Individual sensors on vehicles and equipment will increase real-time intelligence about speed, maintenance needs, temperatures, or other conditions that could cause accidents or damage. There is also tremendous value in integrating data related to traffic, weather and pollution with demographic information and cadastral base maps. Analysis of these combined databases will support proactive decisions, such as issuing alerts to dangerous conditions.
“We are currently experiencing the emerging side of real-time GIS technology and a lot of exciting things are happening,” Joyner says. “Now that we have created the solution, people will be more creative and find new ways to apply it to their problems.”