The Crash Detection Dilemma

I was recently lucky enough to enjoy a track day at the PalmerSport racetrack, which was a gift I’d been given for a significant birthday. It meant I got to drive some fun cars including a Caterham and a Ginetta G56 GTA, which was particularly bonkers!

As an old-school Land Rover Defender driver, I’m not particularly gifted in fast cars, so I spun out several times as I pushed them to the limit, but nothing that seemed particularly serious. So, I was surprised to get a call from my worried wife saying she had received a series of messages telling her I’d been involved in an accident.

Initially, she ignored them, concluding the messages were some sort of spam or phishing attempt. When she eventually called me, I was a little perplexed as I didn’t have my smartphone with me when I was in the cars. I then realized that my Apple Watch Ultra 2 had sent the alerts.

There’s a fascinating dynamic to unpack here. Few notifications are more critical than those alerting that a loved one may have been in an accident or collision. And yet the fact that we both immediately jumped to the conclusion that the messages weren’t real is a testament to the sheer amount of SMS and email spam that exists today and is undoubtedly one of the challenges with these types of service.

It strikes me that there needs to be far greater consumer awareness of these capabilities. At first, I did wonder if the alerts should be clearer, but looking at them again, Apple has gone to great lengths to make them concise and understandable. Advising recipients to call rather than text back makes a lot of sense if someone is in an emergency. Even so, there’s evidently some work to be done here to help people immediately recognize when a notification is a true emergency.

This got me thinking about how crash detection algorithms work on smartphones and wearables, and their implications. There was a tragic case in the UK recently, of a teenage driver whose iPhone alerted the emergency services after a crash, but the alert was ignored and precious minutes were lost. Unlike my false alarm, this was a real emergency.

However, had my smartwatch or phone been configured to call the emergency services, it could have sent an unnecessary alert, as was the case when Apple first launched crash detection on the iPhone 14 and Watch Series 8 and the capability was triggered by passengers on roller-coasters.

This is a potentially huge dilemma for device manufacturers wanting to implement a feature with enormous life-saving potential. If false alarms dramatically outnumber real incidents, the capability will be devalued quickly. It risks undermining public confidence as people may begin to doubt the reliability of the alerts.

Car manufacturers are facing the same challenges as a growing number of vehicles have crash detection systems that alert emergency services or first responders. Huge efforts are being made to improve sensors and detection algorithms, and more steps are being added to the process, such as getting driver confirmation before sending an alert. Of course, if no response is received, the alert is dispatched anyway. Assessing the severity of the crash can also refine the decision to make a crash alert.

The advantage vehicle manufacturers have is their ability to use the sensor framework on the vehicle itself, which is something smartphone and smartwatch manufacturers can’t yet access.

In my case, I think it’s a relatively easy fix. Motorsports organizers should suggest people not only leave their phones out of the cars, but also take off smartwatches if they have crash detection capability.

It also struck me that there needs to be greater awareness of this functionality, and emergency services need clear procedures to deal with it. This is especially important as I expect a growing number of connected devices — be they smartphones, smartwatches or cars — will have increasingly sophisticated crash detection systems that will trigger alerts.

For me, it’s been fascinating seeing this first-hand. It’s astounding to see just how capable our devices are in looking out for our safety and well-being, and the Apple Watch deserves plaudits for identifying that I did, in fact, spin out in a car. But the boundaries between incident and alert need to be managed carefully, and I’m curious to see how device-makers, emergency services, first responders and individuals think about this technology in the future.