The Car That Drives Itself and the Freedom It Returns

Grace Yoon handed her car keys to her son on a Tuesday afternoon fourteen months ago. The accident that prompted the decision was minor and not her fault. A teenager ran a red light in a shopping center parking lot. Nobody was hurt. But Grace sat in the parking lot for twenty minutes afterward and decided she was done. The 2019 Camry went to her son’s driveway. Grace went home.

The decision was reasonable. The consequence has been something else.

She has not been to her cardiologist in four months. The office is open Tuesdays and Thursdays, and those are the days her son works. She has not been to the Korean grocery store in Tempe since November, eight months ago. She has not visited her friend Miriam, who lives eleven miles away, since Christmas. The keys were a car. What Grace gave up was her life at its radius.

What Driving Cessation Actually Costs

#

When researchers study what happens after older adults stop driving, the findings are consistent enough to be uncomfortable. Depression rates increase significantly in the year following driving cessation. Social isolation accelerates. Healthcare utilization drops, not because the person is healthier but because she cannot get to appointments. Falls increase, because the person who cannot drive stops going out, stops walking through parking lots and stores and neighborhoods, and reduced physical activity is one of the most reliable predictors of fall risk.

The cardiologist Grace has not seen in four months is not a detail. It is a health event waiting to happen.

Driving cessation is often framed as a safety decision, which it usually is. It is less often framed as what it actually is: a health event that creates cascading risks that can exceed the risks the decision was made to avoid. Nobody sat down with Grace and said: here is what losing the car will cost you, and here is how we make sure it does not cost you your cardiologist.

What Exists Right Now for Grace

#

The honest assessment of Grace’s options in suburban Phoenix requires separating what exists in theory from what is available at her address.

Rideshare services, Uber and Lyft, cover most of the Phoenix metro area. For a person without a smartphone, GoGoGrandparent operates as a phone-based interface: you call a number, a dispatcher books the ride, the driver arrives. GoGoGrandparent costs around $14 per month in subscription plus the ride fare, and it sends status updates to a family member. It is available now.

Waymo, the autonomous ride-hailing service operated by Alphabet, runs in Phoenix. Grace could, in theory, summon a driverless car to take her to her cardiologist. The practical question is whether Waymo’s current service area includes her specific suburb. Waymo has expanded within Phoenix but still operates in defined geographic zones. If she lives inside the zone, this option is real today. If she lives outside it, it is not.

Non-emergency medical transportation through Medicaid covers qualifying patients for rides to medical appointments. Grace is on Medicare, not Medicaid, which means NEMT does not apply to her unless she has a Medicare Advantage plan with transportation benefits. Checking what her specific plan covers is worth a phone call.

Area Agency on Aging programs in Maricopa County offer volunteer driver coordination for medical appointments. Scheduling requires advance notice, typically three to five business days, and availability depends on volunteer capacity. It is not a same-day solution.

The paratransit system mandated under the Americans with Disabilities Act operates in the Phoenix area. The scheduling requirements and the window of pickup times have been widely criticized as poorly suited to medical appointments with specific times. The system exists. Its reliability is uneven.

Autonomous Vehicles, Honestly

#

Waymo represents the most significant near-term change in transportation options for older adults who cannot or choose not to drive. It is operating in multiple cities, expanding, and works well for users who can manage a smartphone app.

What autonomous ride-hailing cannot do yet matters as much as what it can. Geographic coverage is limited to mapped, maintained urban and suburban environments. Rural areas are not served and will not be served in the near term; the infrastructure requirements, including detailed mapping and well-maintained lane markings, do not exist in most rural geographies. Weather limits performance in heavy rain and snow, which affects cities that have it and largely exempts Phoenix. Accessibility for users with mobility devices varies by vehicle configuration, and the answer to whether a person with a walker or a wheelchair can use a Waymo depends on which vehicle shows up.

The cost question matters on a fixed income. Autonomous ride-hailing is priced similarly to standard rideshare. A round trip to a cardiologist ten miles away might cost $25 to $40. Across twelve months of regular appointments and grocery trips, the transportation cost is meaningful.

The Rural Gap

#

Grace lives in suburban Phoenix, which makes her relatively well positioned for what is coming. Her situation is worth naming in contrast to her friend Margaret in Flagstaff, 150 miles north.

Margaret also stopped driving two years ago. Flagstaff has no rideshare coverage comparable to Phoenix. It has no Waymo. Its paratransit system covers a small geographic area. The volunteer driver program serves the county but has more demand than drivers. Margaret gets to medical appointments when a neighbor is available.

The autonomous vehicle revolution is an urban and suburban phenomenon. The infrastructure it requires, detailed mapping, maintained lanes, commercial density sufficient to make deployment profitable, does not exist in rural America and will not exist within the next five years. The transportation gap between urban and rural aging adults is widening as technology advances, not narrowing. Any honest account of what autonomous vehicles will do for people like Grace has to say that clearly.

What Integration Would Change

#

There is a structural difference between Grace’s current situation and what a coordinated system could provide.

Currently, Grace’s cardiologist needs to be scheduled, transportation needs to be arranged separately, her son needs to be notified, and the return trip requires a second scheduling. Each step is a logistical problem she or her family must solve. When the steps fail, she misses appointments.

The integration that would change this is not science fiction. The personal AI that manages Grace’s health calendar schedules the cardiologist appointment, books the ride to coincide with the appointment time, sends confirmation to her son, and arranges the return pickup. The transportation problem disappears as a separate logistical challenge because it is solved at the same time the appointment is booked, by the same system. That integration is close. It is not fully here.

The difference between “transportation as a problem Grace solves” and “transportation as a component of her health management” is the difference between the four-month gap and the regular appointment.

What It Cannot Replace

#

The ride to the Korean grocery store in Tempe is not only a transportation problem. It is a sensory experience: the smell of the produce, the sound of Korean spoken around her, choosing her own vegetables, the conversation with the owner who has known her for eleven years. An autonomous ride can get Grace to the store. It cannot replace the son who used to drive her and stayed for lunch.

Miriam, eleven miles away, is not an appointment. She is a friend Grace has not seen since Christmas, and the reason for the visit is the visit itself. Transportation can get Grace to Miriam’s door. The relationship that makes the trip worth taking existed before the car and persists after it.

Some of what driving cessation takes is logistical. Technology can address the logistical parts. Some of what it takes is relational. The son who drove Grace to the Korean grocery store was doing two things simultaneously. The ride-hailing service does one.

Grace in Three Years

#

The trajectory is honest and partial. In three years, if autonomous vehicle deployment continues at its current pace and the Phoenix service area expands, Grace will be able to summon a driverless car from her home in suburban Phoenix to her cardiologist’s office, to the Korean grocery store in Tempe, to Miriam’s house eleven miles away. She cannot do all of this today. She can do some of it. The GoGoGrandparent option is available now for everything where cost is manageable and scheduling is not time-sensitive.

What will not be true in three years: the full integration that makes transportation invisible, where the appointment and the ride are a single scheduled event managed by a system that knows her health, her preferences, and her geography. That is coming. It is not here yet.

Grace deserves to know the timeline. The marketing version, that self-driving cars will set seniors free, skips the geography, the cost, the accessibility gaps, and the three-year honest horizon. The real version is that her mobility is beginning to return, unevenly, at a cost, with gaps that the rural reader will feel much longer than Grace will. The keys she gave up fourteen months ago are not coming back. Some of what those keys gave her is coming back in a different form. Not all of it.