Skip to main content

Peru highway speeds up incident detection time with Valerann

Average on Lima Expresa network was 12 minutes - now it's five with data fusion approach
By Adam Hill February 23, 2024 Read time: 2 mins
Traffic in Peru's capital, Lima (© Kirill Neiezhmakov | Dreamstime.com)

Peru’s Lima Expresa, a Vinci concession, says it has reduced detection of incidents on its highways following the introduction last year of Lanternn by Valerann.

The 25km road carries 180,000 vehicles per day, connecting the centre of Peru's capital Lima with the port and airport. 

Speaking on a webinar hosted by IBTTA, Francisco Chenguayen, general manager of Lima Expresa, explained that average incident detection time was 12 minutes, prior to the deployment of Lanternn. 

This could be quicker, he said, if the operator was looking at a monitor on the video wall which picked up the event. But in other cases it could be 20-25 minutes – with perhaps the incident only detected because the emergency services, such as the fire brigade, had arrived.

“Now the average is five minutes and there are events which are detected in even less time," confirmed Chenguayen. "This is helping a lot.”

The Lanternn by Valerann solution uses 'data fusion', bringing together information from a variety of sources such as Here and Waze as well as legacy infrastructure such as loops and cameras, to form a single view of traffic flow through the entire network. 

Michael Vardi, co-founder of Valerann, said: “Once you go to a data-centric approach which says 'bring me all the data sources, give me all the alerts from all the data sources', suddenly you have a complete view of what’s happening on the roads.”

Before, the control room was receiving around 400-450 alerts per week, but that has now doubled to 900-950 per week. With this increase comes some false positives, but Vardi explained that the company works to provide context for the customer, and narrow them down to the ones which require attention.

“Before Lanternn we didn’t know the exact time of the vehicle breaking down or when the accident took place because we had to move the camera around and identify it," Chenguayen confirmed. "So a driver could have been waiting for our support for 10-15 minutes before it was detected. Now we have the certainty that, as soon as something happens, we get that information.”

Related Content

  • Mexican highway concession awarded
    June 13, 2014
    A consortium led by Peru’s Mota-Engil has won the tender for Mexico's Cardel-Poza Rica highway concession, transport and communications ministry (SCT) has announced. The 30-year concession covers construction and operation of a 129 kilometre stretch of highway between Laguna Verde and Gutiérrez Zamora, improving connections in the coastal region of Veracruz state as well as connections to Veracruz port, according to tender documents. Works are expected to complete in 2017.
  • Traffic cameras embrace AI
    December 19, 2022
    Artificial intelligence is spreading into many aspects of mobility – but what about traffic management and enforcement cameras? ITS International invited a few vision experts to ponder a couple of leading questions…
  • New technology revolution in urban traffic control?
    January 26, 2012
    Urban traffic control is a well-defined and practised art. Nevertheless, there are technologies here and on the horizon with the potential to revolutionise how we do things. By Gavin Jackman and Andrew Kirkham, TRL, and Jason Barnes. Distributed monitoring and control of urban traffic networks and flows is nothing new. PC-based Urban Traffic Control (UTC) is now well established and operating in many locations around the world. However, it is worth considering the effects of the huge growth in the use of sm
  • Real-time driving data reveals rush hour congestion on London’s road during tube strike
    February 6, 2017
    Following the warning by London Underground chiefs of tube strikes until lunchtime Wednesday 8 February, Waze, the real-time crowd-sourced sat nav app, issued data collected during the strike on 9 January to show, for the first time, just how badly London's commuters are affected by strike action. According to Waze, on 9th January, data at the peak-time 8.05am showed that 24 per cent of traffic was bumper to bumper– effectively standstill; at this time on a normal day it is usually around 12 per cent. Th