Surtrac – AI enabled traffic signals

1) Sustainability problem: Vehicular idling and congestion in traffic stops. Area: Civic Engagement, Mobility. 

  • Idling in rush-hour traffic  costs the U.S. economy $121 billion a year, mostly due to lost productivity
  • It also produces about 25 billion kilograms of carbon dioxide emissions.
  •  In many urban areas, drivers spend 40 percent of their time idling in traffic.

2)  Technology

  • The Surtrac or Scalable Urban Traffic Control system relies on computerized traffic lights coordinating closely with each other. Radar sensors and cameras at each light detect traffic. Sophisticated AI algorithms use that data to build a timing plan that moves all the vehicles it knows about through the intersection in the most efficient way possible.
  • Each signal then sends the data to traffic intersections downstream so they can plan ahead and can avoid congestion.
  • Surtrac system is interoperable and can use DSRC technology for Vehicle to infrastructure communication, where street signals “talk” and rely data to smart vehicles in real time.





3) Stakeholders

  • City and local governments
  • Department of Transportation
  • Road transport commuters.

4) Deployment 

  • Research cities with highest intersection traffic congestion
  • Partner with the DOT and local governments to install the SURTRAC system in those areas
  • Reach production at scale so as to being down cost of installation of each SURTRAC unit.

JV2610  COMMENT TO ANOTHER BLOG POST (Bacteriophages improve food safety and animal health issues) :

“Antibiotic resistance is one of the major challenges facing the global health community and better alternatives are needed in order to prevent mass causalities from anti-biotic resistant bacteria. The Bacteriophages used in BAFASAL are viruses that need a bacterial cell to replicate. Once they infect a bacterial cell, they quickly replicate using the host cells RNA and other vital proteins and then “lys” or kill the bacteria when the new phages emerge from it. Proteon’s phage technology doesn’t affect the animal’s immune system.”

 UNI – jv2610

AI for surveillance? Is there a cost-effective and ethical solution to using big data for managing social misconduct?

Sustainability problem– Rising rates of social misconduct, ranging from littering and improper waste management to crime and violence against women.

While these may not appear to be a “sustainability” problem, we must acknowledge that sustainability refers to the environmental, social and governance aspects of society. Improper social conduct can result in ripple effects that can compound and impact the sustainability performance of a city.

Sustainability technology– Artificial intelligence to convert CCTVs from “solving” to “preventing” social misconduct

Chicago has recently piloted a program where the police use artificial intelligence algorithms to rate every person’s arrest with a numerical threat score. This algorithm shapes policing strategy, the use of force, and threatens to alter suspicion on the streets. In practical effect, the personalized threat score automatically displays on police computer dashboards so an officer can know the relative risk of the suspect being stopped. The predictive score also shapes who gets targeted for proactive police intervention.

This use of big data and machine learning can be viewed as both a terrific advancement and a terrifying example of social control (A popular Japanese anime called Psycho Pass plays with this concept and depicts how society is ultimately controlled by a massive AI system that dictates how people should behave, what is a crime and how the police should handle the situation). However, the threat can be tackled in the following ways-

  • Ensuring transparency- the variables that go into computing the threat score and the logic behind predictions should be known to the public
  • Law makers must ensure that final decisions being made are at the police officer’s discretion- the final authority on a decision should always be a human
  • Ensure that there is a clear distinction between the algorithmic output and human decisions and bias

However, the purpose of this post is to introduce another interesting way of using AI to monitor social conduct that is less intrusive. An example of this can be seen in the 24/7 surveillance model piloted by Kolkatta in India in 2012 (and subsequently replicated in a few cities globally). The system relies on existing infrastructure (CCTV cameras installed across cities) and uses video analytics and artificial intelligence algorithms to identify anomalies in behaviour.

Picture a normal every day scene on a busy pavement. Office-goers on their way to work, pedestrians grabbing a bite at a food truck. The moment something out of the ordinary happens – someone lunges at another person, a pedestrian collapses, a crowd suddenly gathers or a bag is left unattended too long – intelligent algorithms will instantly identify any change in the normal picture and alert a computer placed in the nearest police kiosk, which will set off an alarm at the local police station through satellite connectivity. In just a few seconds of a suspicious activity or object being detected, officers will be watching it live on their screens and initiate appropriate action.

The technology can be used to ensure corrective action and adherence to rules regarding traffic, waste management and littering as well as more serious crimes. However, the exact placement of CCTV cameras is a sensitive issues- crowded streets and public areas are a given while private buildings and residences will always be out of bounds to ensure respect of privacy.


Key stakeholders and their role in implementation

  • Governments- to ensure that proper and detailed guidance on the ethics of using such systems are in place along with ensuring protection of human rights
  • Enforcement agencies- including the crime and traffic department of police, Department of Sanitation etc.
  • Citizens- to voice their opinions, understand the terms and conditions and ensure they are contributing to the formulation of guiding policies


Post on Solar Bike Paths-

This is a fantastic thought and a perfect use of existing space. I myself have thought of alternatives to such innovations- how about mounting small solar modules on the top of buses and trucks that spend hours on end out on the road in the sun?

The issue i see is with grid management- will this solution exacerbate the duck curve problem in times of over generation? I think that for this to reach scale, we need to constantly think of storage at scale as well.

-By Aksheya Chandar (ac4154)