Time to Collision (TTC) is a widely deployed safety indicator used to estimate the time it would take two vehicles (or a vehicle and an obstacle) to collide if they each maintain their current speed and trajectory. TTC has been used in the vehicle safety context for many decades since US engineer John C. Hayward introduced it in 1972. Today, it's commonly employed in driver assistance systems, such as collision warning systems.

TTC provides a simple, intuitive way to estimate the collision risk between two vehicles. Its quantitative measure of safety allows for the comparison of different driving scenarios or the evaluation of how effective different safety interventions are.

However, TTC has significant limitations. TTC is sensitive to small changes in vehicle speed or trajectory, which can lead to substantial changes in the estimated time to collision, which in turn can result in false alarms or missed warnings in collision warning systems, reducing their effectiveness.

Another limitation: TTC assumes that both vehicles will maintain their current speed and trajectory, which is rarely the case in real-world driving situations. In addition, TTC does not consider other factors that can influence the risk of collision, such as driver behavior, road conditions, or visibility.

Perhaps most importantly, TTC does not consider the severity of the potential incident. When the driver collides, TTC is zero regardless of whether Vehicle 1 barely nudges Vehicle 2 with a small differential velocity, or slams into it with a differential velocity of 70 mph. 

How can Traffic Safety and Mobility professionals navigate these constraints in the new domain of automated and autonomous driving? 

We have created SHM™, the Streetscope Collision Hazard Measure, our context-sensitive collision measure that surpasses TTC and other widely used, context-free approaches (such as cellphone-based IMU data). 

SHM™ overcomes the limitations of existing safety gauges, provides a leading independent indication of safety, and does not require assumptions or predictions of behaviors. All this in addition to the main difference of SHM™ with current practice: our freedom from trailing data as the baseline. 

We highly recommend reading our seminal Whitepaper, authored by our Co-Founder and CTO, Erik Antonsson here at arXiv.

And then, let’s talk about putting this in use for your Safety Case efforts. Let us know.

‍

\\\

‍

Photo by Yasser Abu-Ghdaib on Unsplash

‍