Skip to main content

AI • IOT • NETWORK EDGE

Why Mass Transit Needs Streaming Analytics

A United Nations report predicts that 66 percent of the global population will inhabit urban areas by 2050. As the world becomes more urbanized, more demands will be placed on mass transit.

In response, municipalities and transit systems are turning to the IoT and big data. Real-time vehicle tracking can optimize routes and minimize overcrowding. Predictive maintenance can minimize breakdowns. Transportation authorities can even manage congestion by adjusting pricing as demand ebbs and flows.

To acquire the data needed for these applications, transit systems must be rethought to support a connected vehicle architecture. The three key resources required for this transition are familiar to the transportation and logistics industry: passengers and cargo, crew members, and vehicles.

Transportation Architectures of Today and Tomorrow

In transit systems, these resources combine to form a "trip", or commercial journey from point A to point B. Trips are supported by logistics processes that manage these resources (Figure 1). All of these concepts together represent a transportation network.

Figure 1. A connected vehicle architecture incorporates passengers/cargo, crew, vehicles, and processes. (Source: TIBCO Software Inc.)

While this model is not new, traditional architectures have a major shortcoming: They do not efficiently share information between resources and systems.

The resulting data silos can present challenges, as each resource is dependent on the others. For example, a shortage of maintenance and engineering personnel can result in fewer operational vehicles and less passenger capacity.

But today the IoT allows data to be quickly sourced from multiple transit resources using a variety of sensors and communications technologies. It also provides a foundation for streaming analytics and smart event processing. All that's needed is a unified platform that aggregates data from different resources in one place.

Liberating Connected Vehicle Data

One such platform is the Connected Vehicles Accelerator (CVA) from TIBCO Software Inc. Using a suite of TIBCO software technologies, the CVA gathers data from the resources, processes, and business systems that make up a transportation network. The ingested information is then normalized using the general transit feed specification (GTFS) and used to create a data model within the accelerator.

TIBCO software components that contribute to the CVA include (Figure 2):

  • TIBCO Enterprise Message Service messaging middleware, which supports the integration of heterogeneous platforms and is compatible with Java, Java Messaging Service (JMS), C, .NET, CICS, and COBOL libraries
  • TIBCO ActiveSpaces, an in-memory object storage solution that allows multiple applications to concurrently read and write into low-latency data grids
  • TIBCO StreamBase analyzes and acts on IoT and other streaming data in real time
  • TIBCO BusinessEvents enables people, systems, and devices to interact in real time
  • TIBCO Live Datamart performs continuous queries and computations against high-speed streaming data and events.

Figure 2. The TIBCO Connected Vehicle Accelerator (CVA) encompasses a suite of software components. (Source: TIBCO Software Inc.)

Once information has been captured in a CVA data model, rules can be applied to detect events that deviate from the norm and alert certain stakeholders in real time (Figure 3). For instance, GPS data could be used to alert passengers and crew of a train running behind schedule. Or an automated action could reroute a nearby train or bus if the scheduled vehicle breaks down.

Figure 3. The TIBCO Connected Vehicle Accelerator (CVA) aggregates transit system data into a central repository capable of issuing real-time alerts, actions, and advisories. (Source: TIBCO Software Inc.)

Where other IoT implementations depend on layers of analytics processing, the TIBCO ActiveSpaces component of the CVA uses an in-memory model that allows data to be stored in RAM or flash as opposed to disk storage. This permits much faster data access times, eliminates seek time associated with relational databases, requires simpler algorithms, and executes fewer CPU instructions. As a result, streaming data from all of the resources, processes, and business systems in a transportation network can be analyzed and processed instantaneously.

Analyzed data is also immediately fed back into the CVA data model so that the network self-optimizes over time. Operators can monitor this progression using a real-time operations dashboard supported by TIBCO Live Datamart.

A Connected Vehicle Case Study

Dutch Railways is the principal commuter railway operator in the Netherlands, serving 1.1 million passengers per day on 4,800 scheduled domestic trips. The company operates and manages more than 800 trains.

Each train in Dutch Railways' transportation network has its own onboard information systems that produce roughly 50 points of telemetry data per second. In essence, each train is a mobile data center that constantly generates real-time data. But in the past that data was siloed from the main Dutch Railways data center, which houses the back-end business systems that support the rail network.

The company realized that its transportation network could benefit from what it calls a virtual train. The virtual train is a digital asset that combines data from a train's onboard information systems and Dutch Railways' back-end data center to provide better insights for passengers and crew.

Dutch Railways achieved this by implementing the TIBCO CVA in its main data center, as well as in the onboard information systems of each train. In both cases the TIBCO software runs on Intel® processors, which provide ample integrated RAM and flash to support the in-memory data model, streaming analytics, and event processing.

Now information from virtual trains is accessible by passengers and crew members. This information ranges from a real-time map of train locations throughout the country all the way down to which seats are presently occupied on a given train.

Dutch Railways has also begun connecting front cameras on its trains to the CVA, which could help identify railway obstructions or damaged tracks. Depending on the situation, an automated action could be prompted to stop the train or dispatch crew for repairs.

Connected Vehicles: From Data Silos to Data Sharing

Just as Dutch Railways exposed real-time information to passengers and crew members, data from a connected vehicle architecture can be shared with partners through a set of standard APIs. This allows real-time data to be put to use in a number of ways, such as sharing vehicle telemetry data with a maintenance and engineering firm or marketing passenger information to third parties.

Despite its name, the connected vehicle is less about communications and more about distributing data to stakeholders as quickly as possible. With support from off-the-shelf components, new transportation network architectures enable this in real time.

About the Author

Brandon is a long-time contributor to insight.tech going back to its days as Embedded Innovator, with more than a decade of high-tech journalism and media experience in previous roles as Editor-in-Chief of electronics engineering publication Embedded Computing Design, co-host of the Embedded Insiders podcast, and co-chair of live and virtual events such as Industrial IoT University at Sensors Expo and the IoT Device Security Conference. Brandon currently serves as marketing officer for electronic hardware standards organization, PICMG, where he helps evangelize the use of open standards-based technology. Brandon’s coverage focuses on artificial intelligence and machine learning, the Internet of Things, cybersecurity, embedded processors, edge computing, prototyping kits, and safety-critical systems, but extends to any topic of interest to the electronic design community. Drop him a line at techielew@gmail.com, DM him on Twitter @techielew, or connect with him on LinkedIn.

Profile Photo of Brandon Lewis