• November 7, 2019
  • 8 minutes
  • 0

Five ways cities are using data and tech to curb violence

From studying women's mobility patterns to listening for gunshots

The woman’s hand shakes as she picks up her phone. Her breathing, like her thinking, is rapid and panicked. What can she do? She opens a new app on her phone, and works through its breathing exercises to calm herself. Then, it tells her what her options are.

As cities grow in size and complexity, governments are taking advantage of advances in data and technology to reduce violence – of all types – to make urban areas safer for citizens.

While apps and algorithms alone can’t prevent or respond to all incidents of violence, the initiatives listed below show examples of how such tools can inform policymakers and citizens alike.

Santiago: Gendered mobility

Researchers in Chile’s capital studied the anonymised mobile phone data of around 250,000 women for three months — the largest gender mobility sample ever. They showed clearly how women have very different and more restricted travel patterns than men. This difference is even greater for poorer women.

Santiago is “somewhat representative” of the wider Latin American region, with socio-economic inequality but also a well-established mobility plan, according to Stefaan Verhulst, co-founder and chief research and development officer of the Governance Laboratory, New York University.

Safety was not the only factor influencing women’s travel, said Verhulst, there were also economic and familial reasons for why they moved around as they did. But the knowledge the study gained can be used to sensitise policy development to gender differences.

In particular, the study found women travelled to hospitals and taxi stands far more than men.

According to Verhulst, the gendered policy implications of this are: “What is the accessibility to hospitals from public transport? How can we design taxi stands in a manner that enhances access, but also safety for women?”

While the study conclusively proved women held different mobility patterns to mean, further research is required to identify what behaviours are specifically influenced by safety concerns.

Mexico City: Have my Back

In Mexico City, 65% of women who use public transport have experienced sexual harassment moving around the city.

While the metro has designated female-only carriages, World Bank researchers Karla Dominguez Gonzalez and Bianca Bianchi Alves were concerned these were only a temporary solution and even risked worsening the problem by reinforcing stereotypes.

Their focus group work indicated there was often little bystander intervention during harassment because the people around feared for their safety. Gonzalez and Alvez piloted an intervention to enable bystanders to interrupt harassment without confrontation.

Their team developed a mobile application for a pilot intervention which allowed passengers to discreetly make a report if they experienced or witnessed harassment on buses. These reports triggered an escalating response, including a pre-recorded announcement on the bus to publicly shame harassers, stopping the bus, and alerting the police.

There was also a marketing campaign, to give passengers ideas for how to act in the event of harassment, and bus drivers were also given specialised training.

An evaluation by George Washington University found bystanders were more willing to intervene the longer they were exposed to the campaign, and risk perception among passengers improved. The researchers also recommended running longer campaigns, specifically targeting men and youngsters.

Tirana: Find Your Voice

An app to help domestic violence victims in Albania was produced by three teenage girls for a competition, run by Technovation, a non-profit working with young women to develop their technology and problem-solving skills.

Having previously learned to code at a US embassy workshop, Arla Hoxha, Dea Rrozhani and Jonada Shukarasi, all 16-years-old, developed GjejZâ. Its tagline was “Find your Voice”, after research showed half of all Albanian women suffered abuse in 2018.

Working with psychologists and women’s rights experts, GjejZâ was designed to help women identify if they were being abused, and empower and enable them to action.

The app’s scope is broad. Features include a series of questions about the woman’s circumstances, testimonies and encouragement from other domestic abuse survivors and even breathing exercises. Others are more practically focused and can connect women with relevant state officials or direct them to shelters or work opportunities.

US: Shotspotter

Audio sensors designed to pick up the sound of gunshots are mounted throughout cities and report violent events using firearms directly to a control centre and then law enforcement. Bypassing 911 calls, the technology speeds up emergency responses — the location of gunfire is triangulated by the sensors – and alerts officers to incidents that might not otherwise be reported. Shotpotter analysts in the control centre can also alert police to the type of weapon used and other crime scene information.

Shotspotter, founded in 1996, is a public company working with local governments and police to introduce the gunfire detection software to over 100 cities across the US. It was also hailed as a “game-changer” by executives when it was installed in South Africa’s Kruger National Park to use against poaching.

The technology has been praised by some politicians and police departments but also come under fire for its cost, with each square mile covered costing around $60,000 annually.

Rio de Janeiro: Crime Forecasting

Igarapé Institute, a Brazil-based think tank, first developed Crime Radar to monitor and predict crime using artificial intelligence during the 2016 Olympics.

It worked on the basis that most crime is restricted to a small number of areas — dubbed “hotspots” — and at relatively predictable times. Information on past murders, police killings, violent attacks and robberies were fed into the platform.

Crime Radar uses an algorithm to collate this historical data and also process the new information it is regularly updated with. In addition to visualising this data by mapping crime hotspots, and safer areas, in real-time, the platform’s machine learning enables it to predict crimes in certain areas.

The risk of crime in a neighbourhood is displayed from green (low risk) to red (high risk) and user can inspect how this is likely to change through the day or week.

Pilots of an updated version are ongoing across the Brazilian state of Santa Catarina, and an evaluation of its effectiveness – particularly on planning police patrols and operations – will be evaluated in 2020. — Will Worley

(Picture credit: Pixabay)


Leave a Reply

to leave a comment.
Master the skills you need for the public service.

Discover inspiring resources, tools and policies.