There are certain neighborhoods that you just can’t walk around in at night. It’s hard to explain, but the only thing you know is that being there leaves you with a creeping, niggling feeling. How do you report this feeling without sounding crazy? What makes people feel safe when they’re in a neighborhood? Wide streets? Bright lights? A certain number of cars?
The two founders of Safetipin, Ashish Basu and Kalpana Vishwanath, have narrowed safety down to nine factors: lighting, openness, visibility, people density, security, walk path, transportation in the area, and feeling.
Safetipin crowdsources information on these nine factors in order to paint an accurate and transparent picture of cities.
The first eight work as independent variables, or factors that are measured on their own in the physical world. The last, “feeling”, counts as a dependent variable and is observed in conjunction with the first eight. “To explain a tangible like “feeling”, you need backup data,” explains Ashish. “We use the preceding data to understand changes in it.”
Once a certain number of these audits are conducted in an area, they are aggregated to create a “safety score”. Safety scores range from one to ten and are represented in that order by green, amber, or red pins. Audits can be shared on social networks and users can create “walls” about certain areas that will then allow them to post info online like broken traffic lights and bad roads.
These are all accessible on Safetipin’s app, which also provides access to information about amenities like 24-hour pharmacies and ATMs. Among other things, the app also acts as a GPS tracker. “It’s over-configured,” Ashish jokes. The app is available in English, Hindi, Bahasa, Spanish, and Mandarin and has had over 40,000 contributions. It eventually hopes to cover 100 cities across South Asia, Southeast Asia, Africa, and South America and currently works in one city in each of these areas: Jakarta, Bogota, Nairobi, and Delhi.
The difficult thing about crowdsourcing information is maintaining quantity and quality.
In India, Safetipin has been launched in what is known as the country’s “rape capital”, Delhi. While that’s a horrible and degrading way to describe a city – as if rapes can be quantified and addressed in the same way that a certain number of startups can define a “startup capital” or a proliferation of food stalls can define a “street food capital” – the number of reported rapes in Delhi are more than anywhere else in the country. Reported cases doubled from 706 to 1,638 between 2012 and 2013. Last year saw 2,166 reported rape cases.
However, it is important to remember that this does not necessarily mean that the number of rapes are increasing. In fact, it might actually be a positive thing – perhaps, more people are reporting rape cases than ever before.
“More data is better data,” Ashish explains. “That was something important that we had to learn. You can’t have a sample size of two people telling you that an area is unsafe. You’re looking for more information.”
Because it can be unattractive for a business to exist in an unsafe area, some people go so far as to lie on these services. In order to address this problem, Ashish explains that the amount of data that the app has is important. “If you look at it in pure mathematical terms, those who fabricate numbers will be outliers.”
Of course, it’s not always easy to get these numbers. Particularly in apps that rely on social interaction like Safetipin, it can be difficult to motivate users to contribute. “The interesting trend was that many people would download the app but would then stop at one or two audits,” reminisces Ashish. “We tried everything from target groups to send out a rubric based set of questions. Eventually we found a bunch of volunteers who helped us crowdsource information.” With more contribution and change, however, the app has picked up. “That’s how behavior works on social networks,” says Ashish. “Once the ball gets rolling, it takes a lot to stop it.”