Making cycling safer through AI and machine learning

Cycling fatalities can be prevented through a number of schemes which incorporate machine learning and artificial intelligence.

The cycling revolution that has swept the western world shows no signs of slowing down with record levels of people getting on their bikes in the United States and the UK in recent years. But with rising numbers comes an increased concern about the safety of cyclists – particularly in large cities.

In fact, cities such as New York are seeing more deaths than ever from cycling while three cyclists were killed in just four days in London in February 2017.

In addition, concerns over a growing sedentary lifestyle and soaring obesity rates necessitate bike share schemes which are easy to access within cities.

Fortunately, the boom in artificial intelligence and machine learning companies have provided new innovations when it comes to making our journeys safer for those who prefer to commute to work on two wheels instead of four.

Machine learning: giving cycling a head start

Machine learning could make urban cycling a whole lot safer through a phenomenon frequently sought by motorists and cyclists akin to the perfect wave for surfers.

The “green wave” occurs when a commuter hits a sequence of green lights ensuring they do not need to stop. While incredibly satisfying and time efficient, this occurrence could also improve cycling conditions.

Traffic light-mounted cameras can detect approaching cyclists and analyze their actions in conjunction with deep learning algorithms. Machine learning ensures the lights can recognize cyclists in variable conditions. The implementation of Graphic Processing Units into embedded computers ensures it can react in real time to oncoming traffic. Following a successful trial in Germany last year, the prospect of safer cycling conditions seems closer than ever.

Driverless cars: helping drivers and cyclists

Masked by all the excitement around the potential of artificial intelligence in creating driverless cars is the increased safety we can expect for cyclists.

The self-driving car project Waymo is programming their vehicles to recognize certain cycling tendencies. For example, they are training cars to predict a cyclist’s course by analyzing their hand signals before they make a turn or change lanes.

Researchers are programming driverless cars to examine their surroundings in order to be safer around bicycles. If a parked car has a door open, cyclists will move further out in order to avoid hitting it. Autonomous vehicles built on AI will be able to detect these situations and cater for them by giving way beforehand in order to provide the necessary space to avoid collisions.

Artificial intelligence: a growing role in city bike share

With more than 68% of adults in the United States considered obese and health experts recommending a minimum of 2.5 hours of aerobic exercise per week, cycling to and from work is one of the most efficient ways to meet this requirement and improve the population’s health. Therefore, it is imperative that bike share schemes make their bicycles as easy to access as possible during the hustle and bustle of city life.

Given the amount of data created in bike share schemes, AI can be used to drive that efficiency. Companies such as General Bikeshare Feed Specification are now able to piece together information about peak times, the most popular areas and any areas where they can scale back supply.

An investment in AI for bike share schemes will provide cost benefits to companies who will not need to maintain such a large fleet to provide their service. In addition, a well-run scheme will increase participation among commuters.

The rise of pedal power

In the center of London, the ratio of cars to cyclists has increased from 11 to 1 in 2000 to just 1.7 cars for each cyclist in 2014. This explosion in demand has focused the spotlight on cycling safety in some of the world’s most prosperous cities.

Fortunately, this growing concern could be solved sooner than we think. We have already seen accounts of new AI technologies such as driverless cars showing an awareness of cycle safety. Once the stabilizers are off, these solutions will prove as valuable to the protection of cyclists as the helmet.

Inbenta is a leader in natural language processing and artificial intelligence for customer support, e-commerce and conversational chatbots, providing an easy-to-deploy solution that improves customer satisfaction, reduces support costs, and increases revenue.

Interested in finding out more? Our team of experts is available to show you how Inbenta can benefit your company.

Let’s get in touch