Cognitive Technologies to develop autonomous tram in Russia

Cognitive Technologies has joined forces with Russian manufacturer PC Transport Systems to deploy an autonomous tram on the streets of Moscow by 2022. Cognitive says that its simplified system means autonomous trams will appear on public roads much earlier than self-driving cars. The company claims its system will detect vehicle and other trams, traffic lights, pedestrians, tram and bus stops, railway and switches and obstacles. Also, the technology will allow the tram to stop in front of obstacles a
February 14, 2019 Read time: 2 mins

Cognitive Technologies has joined forces with Russian manufacturer PC Transport Systems to deploy an autonomous tram on the streets of Moscow by 2022.

Cognitive says that its simplified system means autonomous trams will appear on public roads much earlier than self-driving cars.
 
The company claims its system will detect vehicle and other trams, traffic lights, pedestrians, tram and bus stops, railway and switches and obstacles. Also, the technology will allow the tram to stop in front of obstacles and maintain a safe distance to the cars ahead, accelerate and stop.

The trams will feature a combination of sensors which include 20 video cameras and up to ten radars to help detect road scene objects at night as well as in rain, fog and snowy conditions.
 
Olga Uskova, president of Cognitive, says the company’s low-level data fusion technology allows the computer vision model to use the combined raw data coming from cameras and radars to provide a better understanding of the road scene.

“Cameras, for example, correctly recognise objects in 80% of cases, additional data from radar raises the detection accuracy to 99% and higher,” Uskova adds.

The trams will use GPS sensors and will use high-precision cartography along its route.

Initially, an intelligent control system will serve as an active driving assistant in dangerous situations. A second stage test will follow in which an operator will remain in the cabin as a backup driver.

During the next two months, autonomous tram tests with the operator in the cabin will take place in closed facilities which will then be followed by a trail in Moscow.

Related Content

  • Shock therapy: jolt for EV charging needed
    October 2, 2018
    As sales of electric vehicles accelerate, the growth of charging infrastructure is in need of a big boost. Graham Anderson reports on whether Europe is up to it. Utilities, technology companies and vehicle manufacturers are battling to put in place new charging networks for electric vehicles (EVs) across Europe in response to a predicted dramatic surge in demand. Market experts believe that rapidly falling battery costs – which make up about one third of the costs of an electric car – and growing
    Read More
  • Switching Atlanta onto MaaS
    May 9, 2019
    It’s easy to talk about MaaS in the abstract – but MaaS isn’t going to work if it’s just a theory. Colin Sowman speaks to one woman about the practical benefits - and difficulties - of getting out of her car and switching to public transit in Atlanta, Georgia One of the first goals of Mobility as a Service (MaaS) inventor Sampo Hietanen is that MaaS should persuade households they don’t need a second car. This is starting to happen - even in the car-dominated US. Last year, authorities in the state of Ge
    Read More
  • Machine vision makes red light enforcement easier
    December 1, 2015
    Teledyne Dalsa’s Manny Romero looks at how the combination of camera manufacturer and software provider can make enforcement easier. Californian video analytics solution provider Eutecus develops real-time images capture and high speeds processing technology for applications including intelligent lighting and advanced driver assistance systems.
    Read More
  • Hackers can fool self-driving car sensors into evasive action
    September 8, 2015
    The laser ranging (LIDAR) systems that most self-driving cars rely on to sense obstacles can be hacked by a setup costing just US$60, a security researcher has told IEEE spectrum. According to Jonathan Petit, principal scientist at software security company Security Innovation, he can take echoes of a fake car, pedestrian or wall and put them in any location. Using such a system, which he designed using a low-power laser and pulse generator, attackers could trick a self-driving car into thinking somethin
    Read More

Highlighted Content