Autonomy in Motion: Enabling Autonomous Applications for Transit

Aug. 15, 2018
Achieving true autonomy and automation requires a mobile wireless network and real-time data.

“The techniques of automation are already permitting us to do many things we simply could not do otherwise. Some of our largest industries, some of our largest employers would not exist and could not operate without automation ... it can remove fullness from the work of man and provide him with more than man has ever had before.”

Quick, name that quote. Is it from a TED Talk? Is it from a presentation at TechCrunch Disrupt or the Future Festival? Did Elon Musk say it?

None of the above. Those words were uttered by President Lyndon B. Johnson more than 50 years ago, after he signed the 1964 bill creating the National Commission on Technology, Automation, and Economic Progress. While the words are old, the sentiments remain fresh — and the technology enabling automation and autonomy has advanced by leaps and bounds.

Industries of all types now regularly incorporate levels of automated and autonomous technologies, including the transit and transportation industry. Finland, Singapore and China are testing autonomous buses; cities already use artificial intelligence to predict and detect traffic conditions and accidents; and GE’s “intelligent” locomotives aim to import the efficiency of rail transport, according to Tech Emergence.

Yet challenges remain before we can take the next step into fully autonomous trains and buses — not least of which is the need to ensure that transit vehicles remain constantly connected through real-time data. Without a constant flow of up-to-the-millisecond-accurate data, true autonomy cannot be achieved — and this means the need for ironclad network infrastructure is critical to ensuring autonomous applications operate flawlessly.

What autonomy needs

Autonomous applications and automation allow transit vehicles/equipment to operate independently of humans, using sensors, software, computers and network interconnectivity to give the machines a more immersive existence in its environment.

Some applications, such as those controlling certain hazardous train crossings, may require a remote operator, while others will be completely human-free. But for any autonomous application, a robust network is required to ensure the applications and tasks run safely and seamlessly. A delay in transmitting data between autonomous vehicles and a command center, or an unexpected interruption in communications, can disrupt operations. This is true even when the interruption is measured in milliseconds – and this has serious implications on efficiency and safety.

To ensure safety and efficiency are never compromised, some transit agencies are looking to kinetic mesh wireless networks, which enable continuous, mobile connectivity, and can act as the backbone of autonomous applications.

How it works

In a kinetic mesh wireless network, each radio, or node, serves as singular infrastructure, which enables all devices and the network itself to be mobile. It employs multiple radio frequencies and any-node-to-any-node capabilities to continuously and instantly route data via the best available traffic path and frequency, using hundreds of nodes.

If a certain path becomes unavailable for any reason — due to antenna failure or power loss to a piece of equipment, for example — nodes on the network use an alternate route to deliver the data, eliminating any downtime.

The network never fails as a whole; data is sent and received simultaneously, in real time. There are no single points of failure, and the network can be redeployed and scaled in multiple ways simply and easily.

Because there is no central control node, routes are built automatically, and are constantly evaluated for quality and performance, which allows the network to adapt to node location, local interference and congestion dynamically despite conditions that would cripple other networks. It makes the network adaptable, offering a distinct advantage over networks that require all the brain power to reside in controller nodes.

Kinetic mesh in transit

These nodes provide mesh networks with flexibility and stable communications in even the most rugged mobile environments, including transit. These types of networks tend to thrive when applied to systems with moving parts, so to speak. They are well-established in dynamic, rugged industries such as military, mining, and oil and gas.

They also have demonstrated success when deployed on commercial freight trains and are a natural fit for their next venture: mass transit.

This type of network enables radio technology transmission between trains and the wayside to facilitate communication-based train control (CBTC). The nodes located on the trains communicate with all wayside nodes they can see, connecting and passing data between them all, so that when they do drop a wayside connection, there is no disruption — equating to zero data loss onboard the train. As a result, where most technologies have roaming times in the tens of milliseconds, a kinetic mesh network does not have to roam, as it is always connected and passing data between multiple nodes all the time.

Designing each node to serve as infrastructure enables all devices and the network itself to be mobile. Users can employ nodes on all manner of vehicles, including trucks, buses and trains. In addition, these networks can transmit significant amounts of data — 100 Megabits or more, keeping autonomous applications and vehicles running.

A network like this can also be life-saving, in that it can keep track of mission-critical equipment. It can autonomously aggregate data onboard vehicles and send it to the operations center to analyze vehicle state-of-health and other equipment factors. Taking safety a step further into autonomy, the node onboard each vehicle can monitor situations and alert operators that a valve, seal or motor is overheating well before it does, or that a pedestrian is on the train tracks.

Distance can present limitations in a mass transit system, but as long as each node can communicate with another node, the network will preserve its integrity and connectivity will never fail. The key is in the design. When designing mesh networks for mass transit, technical staff account for variables unique to their industry, such as tunnels, subways, train dimensions, node-to-antenna ratios, and size of antennas.

All of these network qualities translate into constant access to real-time data and unfailing connectivity, providing the strong backbone that autonomous applications and automation will need.

A fully autonomous future

All the autonomous applications in the world are for naught if they’re not supported by a network that ensures constant connectivity and unfailing access to real-time data — especially in a transit environment, where efficiency and safety are dually important.

With kinetic mesh networks, every node can be mobile, providing every transportation asset that traverses or resides in the network with robust, high-bandwidth mobile connectivity. This means that command centers can instantly access the analytical insights from applications running on each train or bus — monitoring everything from equipment health to speed and location, wheel sensors and freight status, fuel consumption, traffic optimization and more.

Autonomous applications and automation are ushering in a new age of smart, connected transit, but achieving this requires highly reliable, mobile, resilient communications networks that run continuously and ensure data is never delayed or not received. While we haven’t reached true autonomy yet, establishing a scalable mobile network that can expand to support future applications will put transit agencies one step closer.

Peter Lenard is senior vice president of business development for Rajant.