Dim Traffic Sensors Dull How ‘Smart’ Freeways Are

California’s highways aren’t as smart as they used to be.

Buried under thousands of miles of pavement are 27,000 traffic sensors that are supposed to help troubleshoot both daily commutes and long-term maintenance needs on some of the nation’s most heavily used and congested roadways. And about 9,000 of them do not work.

The sensors are a key part of the “intelligent transportation” system designed, for example, to detect the congestion that quickly builds before crews can get out and clear an accident.

A speedy response matters: Every minute a lane is blocked during rush hour means about four extra minutes of traffic. Fewer sensors can mean slower response times, so the fact that 34 percent are offline – up from 26 percent in 2009 – creates an extra headache in California’s already-sickly traffic situation.

“(It) is not an acceptable number, really,” said California’s top transportation official, Brian Kelly.

With limited space and money for new lanes, Kelly said, maximizing flow on existing freeways is critical. To do so, planners rely on a network of cameras, above-road detectors, message boards and the in-road sensors called “loops” because of their shape.

Some loops were cut during construction, others yanked out by copper wire thieves. Many have succumbed to old age.

The resulting blind spots show up as strings of gray amid the green, yellow or red on the large map that freeway managers overseeing Los Angeles and Ventura counties monitor for signs of trouble. Even worse off than LA, according to Caltrans, are inland areas such as the San Joaquin Valley and San Bernardino and Riverside counties.

The outages are significant enough that the sensors alone cannot produce real-time traffic maps that are useful to the public. Especially when compared to the many private traffic mapping services that drivers rely on to get around.

So, to post online traffic maps that are ready for public consumption, California and other states are paying the private sector.

Caltrans gives away data from its working loop sensors to Google and other companies; Caltrans also pays Google for a traffic map that incorporates its own data as well as information the tech giant gets from vehicles and cellphones whose owners have agreed to share location data.

California’s tab is not large – Caltrans estimates it at $25,000 per year for its public-facing Quickmap – but other states are giving away sensor data and buying back reliable maps as well. Michigan’s transportation department said it pays Inrix Inc. about $400,000 annually for data to populate its Mi Drive map.

An Inrix spokesman said the company has contracts with 25 state transportation departments.

Loops are a simple technology that can last decades when properly installed. A bundle of wires under the pavement detects the size, speed and number of vehicles that pass over it, transmitting the information to a roadside box. That data records traffic in real time, but also helps planners who want to know how many of what kinds of vehicles use a road so they can project when it will start to deteriorate (more big trucks means more potholes, sooner).

Drivers may be familiar with loops at surface street intersections, where a circular cut in a turn lane means a loop will detect an idling car and tell the light to change. Replacement materials cost only a few hundred dollars – but installing a loop on a freeway can cost thousands because to embed the wire crews must close two lanes, likely off hours when labor is more expensive.

In the Fresno and San Francisco Bay areas alone, Caltrans plans to spend $35 million to fix loops sensors – as well as freeway lights, cameras, ramp meters and other electrical systems – that are down due to metal scavengers or other problems.

The state that pioneered the use of loop sensors starting in the 1970s is not alone in its struggle to keep them producing reliable data.

In Utah, transportation officials estimated about 20 percent of loops do not work.

“Does it impair our ability to make informed decisions? Certainly,” said Blaine Leonard, manager of the state’s intelligent transportation systems program.

Information from loops informs the estimated travel times posted on freeway message boards.

“If the data is bad and therefore the travel times are bad, at some point in time the public goes, ‘Well, they don’t know what they’re doing,”‘ Leonard said.

About 75 percent of loops In the Austin, Texas area are not working due to large-scale freeway resurfacing, according to the state department of transportation. Michigan’s transportation planners abandoned loops because they found too many failed during winter’s freeze-thaw cycle; they’ve moved to above-road sensors that use microwaves to detect traffic.