The public outcry over the recent massive recall of potentially dangerous airbags — this one affecting as many as 4.7 million vehicles — is just another reminder that the humble car airbag still has a ways to go before it reliably offers the type of safety protection that Americans have a right to demand. Ever since the first car airbag appeared at the beginning of the 1970s, there have been numerous iterations on the basic design — a giant bag of air that rapidly deploys upon impact — but we simply have not come far enough, fast enough in 40 years.

While it seems as though we are caught in a seemingly endless cycle of incremental innovation for the humble car airbag, there are some out-of-the-box ideas out there. One idea is to supplement airbags inside cars with airbags outside cars in order to protect innocent bystanders. In 2012, for example, Volvo created the world’s first “external” pedestrian airbag for the V40. The aim was to have an airbag that would deploy to prevent pedestrians from being sideswiped by vehicles.

Another idea is to specifically target parts of the human body with airbags that are smaller and deploy with less energy — and thus, avoid some of the problems posed by the Takata airbags involved in the recall. The general idea is to create airbags specifically contoured to parts of the human body. Fifteen years ago, for example, knee airbags were developed to prevent knees from being crushed upon impact with the dashboard. And, back in 2009, Ford Motor Company pioneered “inflatable seat belts” as a way to cushion impact, using a hybrid seat belt-airbag approach to automobile safety.

But the real potential for disruptive innovation in today’s airbag industry is in “pre-crash” airbags that work in concert with sensors and cameras on the vehicle. This would help to reduce one of the most common problems — airbags that deploy upon impact with too much force. Without any advance warning of a crash, airbags must generate enormous force in order to work, and that’s what leads to all sorts of safety problems. In short, it’s like “getting hit in the face with a boxing glove.” With sensors, though, airbags would be able to deploy in advance — and those precious milliseconds might be enough to reduce serious injuries.

Automakers such as Audi and Lexus are now working on “pre-crash” and “pre-collision” systems that include radar sensors and active cameras. If you push hard enough on the concept of sensors and cameras being used to create a “pre-crash” system, you get the self-driving car. That’s where things get really interesting. If you consider that 90 percent of all accidents are caused by driver error and that 40 percent of those accidents can be linked to driver fatigue, alcohol, drugs or other distractions, it’s clear that cars without beer-drinking, texting human drivers — or, at least, cars with enough sensors to detect traffic in front of and in back of them — could have a significant impact on reducing the risk of being involved in a serious car accident.

In 2012, a report from KPMG and the Center for Automotive Research (CAR) in Michigan suggested that the coming era of self-driving vehicles would have a huge number of consequences for automobile safety. They might even change how we build cars. For example, as Professor Raj Rajkumar of Carnegie Mellon University, who is the co-director of the Carnegie Mellon self-driving car project, told me in a phone conversation, self-driving cars would not need to include as much reinforced steel for safety reasons, making them lighter and less dangerous.

And that’s not all. Self-driving cars, if they ever become mainstream, would radically change the car driving experience. There would be fewer traffic jams, fewer traffic lights and less need for road signs. If cars travel together in packs, all traveling at the same speed, as KPMG and CAR suggest, wouldn’t that mean more “fender benders” and less serious types of collisions? You wouldn’t have a mix of aggressive and conservative drivers on the road, you’d have cars all going at about the same speed. As Rajkumar told me, “Speeds would be harmonized and that would help to avoid accidents and traffic jams.”

And, who’s to say that the standard car configuration — a driver and passenger in front, more passengers in back — is even the optimal configuration from a safety perspective? As Rajkumar suggested, “We could see different layouts. The seating could be adaptable, the seats could be turned into more of a roundtable, or the back seat could become a bed.” That would seem to argue for a different approach to airbags that challenges the conventional approach based on frontal airbags for drivers. If cars start to look more like other forms of mass transportation, that begs the obvious question: What other forms of mass transportation have airbags?

The important point is that new innovations in car safety made possible by sensors, radars, lasers and onboard computers could also lead to radically new types of car experiences. And that could lead to a radical re-rethinking of the humble car airbag. If you break it down, the idea of an exploding bag of air to protect drivers is really just a high-tech version of surrounding yourself with pillows. We have a choice: Do we just create a more efficient airbag or do we break open the category entirely and create a new auto safety paradigm in the process? Once the cost of installing sensors in self-driving cars comes down even further, we might see the beginning of the end of the humble car airbag.