As the auto industry finds itself evolving at a clip not seen in decades, entrepreneurs and experts need to look beyond engineering and mechanics to drive real change. The future to which many in this industry aspire will depend just as much on innovation in the key enablers that make the business tick.

That future can be summarized in three points:

• Electric propulsion replacing internal combustion at scale.

• Cars that can drive themselves.

• Vehicles available on-demand; drivers who need a truck one day and a compact the next will be able to summon either option in a matter of minutes.

This would change just about every way we think about, pay for, and use cars, but perhaps nothing more change more than the concept of ownership. Instead of spending thousands of dollars for a depreciating asset, we’d simply pay for what we use. Instead of requiring a person to drive and thus being unavailable to indulge in leisure or get work done, everyone would be a passenger in a self-driving car.

This future depends on quantum changes in just about every aspect of the way cars are designed, built, marketed, sold and used. Given how quickly Uber disrupted the livery business, Amazon upended retail, and airbnb transformed lodging, it’s tempting to imagine the car business changing just as quickly.

Transformation will depend not just on what GM, Ford, Toyota and other manufacturers can do. Radical change must occur within just about every enabler, service and policy that sustains the industry.

Entrepreneurs take note: Here are five enablers that must change to usher in a new age of automotive transportation.

• Insurance: Who will be liable for damages when a self-driving car kills or maims? California requires all self-driving cars to include a steering wheel and a licensed driver behind it, but if people are required should the self-driving feature fail, that wouldn’t be very liberating. People who are too old, too young, too inebriated or otherwise not able to drive would remain shut out.

• Software:

  • It has to become more reliable than any smartphone OS while doing much more than smartphones ever need to do. This will require huge improvements in the way software is made and tested.

  • Providers must be able to update and maintain this software. This means total control over updates without having a manufacturer, operator or even the user get in the way, as is the case with many smartphones.

  • Security: a fleet of self-driving cars must be inter-connected to maximize safety and efficiency. The software that binds it must be impenetrable to hackers, who otherwise could inflict massive damage and loss of life.

• Refueling:

  • Charging stations must be as ubiquitous as gas stations.

  • Charging time must rival the time it takes to fill a 15 gallon gas tank. This will require a massive engineering feat, taking us from about 8 miles a minute for Tesla’s super charging facilities to 30-50 miles per minute while not compromising safety or the integrity of the grid. Here’s a useful article that describes the engineering challenges that stand between the status quo and real innovation.

  • We’d need much more electric power to support rapid charging of a significant portion of the flee. If you’ve ever tripped a circuit breaker in your house, you can imagine the stresses of 20 cars rapidly recharging at a single station.

• Public safety: I suspect the public will resist autonomous ambulances and p0lice cars for fear that they’ll never be able to handle the urgency and stresses of an emergency. A largely self-driving fleet will need to accommodate human drivers.

• Public policy: The federal government, state governments and local municipalities have created tax incentives to encourage the adoption of electric vehicles. These policies are both regressive, transferring thousands of dollars to the wealthy individuals who can afford $40,000 cars, and a strain on the already heavily-burdened public purse. (The top 100 U.S. public pension programs are under-funded today by over $1 trillion.) For these reasons, tax breaks are under constant threat, so the industry will need to find a way to drive pricing down in ways that do not require taxpayer funding.

While there are efforts under way to tackle the issues, solving all of them will take time, money and changes in the public’s attitude. That’s why I think we’re at least 25 years away from this fully-realized future.

In the meantime, how will we know if we’re on track for this better tomorrow? I asked two really smart friends what changes they expect to see in the next five years. Here’s what they had to say:

• Hastings-on-Hudson, my home town of about 8,000, will have three charging stations.

• The best recharging facilities will completely charge electric vehicles in a maximum of three minutes.

• The top range of electric vehicles on a single charge will be 300 miles (Tesla Model S high end gets 295 miles in a single charge. Chevy Bolt and Tesla Model III estimated to get 300 km, or 186 miles, on a single charge. 2016 Volt looks like it will go 111 miles on a single charge.).

I look forward to the follow-up discussion we’ll have in 2021.