Drivers for Uber, Lyft and other ride-hailing services are notoriously unhappy with their earnings — earnings they see diluted by the amount of time they spend waiting for their next rider.

Michael Treasure became intimately familiar with the problem while studying for a doctoral degree focused on on-demand transportation. He found that on average, the driver spent more time waiting for the next ride than with a rider in the car.

Treasure came up with a solution — a patent-pending algorithm that focuses on scheduling rides in advance to match riders with drivers who will be in the same area. Instead of an Uber or Lyft model where the ride is immediately on-demand, Treasure’s new service, Lymousine, allows riders to schedule their rides as soon as 30 minutes in the future.

Uber and many other companies already offer a scheduling feature. Treasure says their models are inefficient.

From Hypepotamus. Story by Holly Beilin.

In about five years, ridesharing has gone from nonexistent to a billion-dollar market, eclipsing taxis in most major markets. It’s projected to double in size by 2020, grow eightfold by 2030, and at the beginning of this year, total ridesharing miles covered equaled a roundtrip to Saturn. It’s safe to say that the industry has blasted off.

But there’s still some ambiguity around the driver side of the equation — namely, around wages. Because the number of riders you pick up can vary by market, time, day, and lots of other factors, pay per hour is also variable. In one of the most cited recent surveys, the average Uber driver made $15.68 per hour, where Lyft drivers made $17.50 per hour. Of course, that’s before factoring in expenses like gas, maintenance, and vehicle depreciation.

An inherent bias has also been observed for drivers — the same study found that hourly earnings tend to deteriorate by age. Younger drivers, who more often pick up the late night, weekend, and surge pricing shifts, make much more.

Michael Treasure became intimately familiar with the issue while pursuing a Doctoral degree focused on the on-demand transportation space. For his own research purposes, he took it upon himself to conduct about 2,500 rides on Uber and Lyft throughout Atlanta and document statistics such as how often the car had passengers and how much time was spent riding around looking for riders.

His data was telling: the average loaded (with a rider) trip miles in a vehicle was 8.87 miles, while the average unloaded trip mile was 4.21 miles, so unloaded relative to loaded miles per trip was approximately 47 percent. The average ride time of a trip with a rider was 18.51 minutes, while the average wait time for the next ride was 23.66 minutes.

In plain terms, Treasure found that on average, the driver spent more time waiting for the next ride than with a rider in the car.

He then attempted to calculate their take-home pay given those statistics. The average day’s revenue for an 11 hour and 38 minutes work day (yes, that’s almost a 12-hour work day!), driving more than 288 miles for the day, would translate to less than $12 per hour in gross revenue — or approximately $0.46/mile.

“These poor results would in part explain why drivers are grossly unhappy with their earnings and why there is such high churn rates in the industry,” says Treasure.

Treasure set about finding a solution to the problem, developing a patent-pending algorithm that focuses on scheduling rides in advance to match riders with drivers that will already be in their area. Instead of an Uber or Lyft model where the ride is immediately on-demand, Treasure’s new service, Lymousine, allows riders to schedule their rides as soon as 30 minutes in the future.

For riders, that means they know exactly when and where they will be picked up — no walking around trying to find your Uber and no surge pricing, even if you’re in a crowded place.

For drivers, it means less of those unloaded miles. No driving around waiting for passengers to pick you up, no guessing where the most business that night might be. Lymousine guarantees drivers a $20 minimum wage per hour.

It might seem redundant, as Uber and many other companies do now offer a scheduling feature. But Treasure says their model is inefficient.

“This option often proves unreliable in practice as drivers simply forget or pick a more profitable or surge rider instead. Many riders thus avoid that in favor of “black cars,” friends, and family when they need to be picked up at a scheduled time,” says Treasure.

Treasure says they aren’t focused on the entire audience of an Uber or Lyft. Passengers stumbling out of a bar who want to go home immediately, for example, are probably not the ideal customer for Lymousine given the required 30-minutes wait time.

But those who need to get to the airport at a certain time? Ideal customers.

The core team of 5 has launched in the Atlanta market with about 15 drivers on the payroll. They intend to double that by year’s end.

In addition to the guaranteed minimum hourly wage, Treasure says 5 percent of a drivers’ earnings are set aside to save for an autonomous vehicle. The impact of self-driving cars was also a portion of his doctoral thesis; Uber and other ridesharing companies have been testing autonomous vehicles for some time, but have yet to explain where the driver fits into the picture.

“Unlike other leading ride sharing companies that intend to get rid of drivers when autonomous vehicles come on line, Lymousine sets aside revenue to help our drivers buy their own autonomous vehicles and continue building a business in a disruptive market rather than being disrupted out of business,” says Treasure.

Finally, the company is using the strength of their algorithm in a B2B play as well. For retailers looking to compete with the likes of an e-commerce giant like Amazon, Lymousine offers a last-mile, same-day package delivery solution.

“Consumers want convenient same day and instant package delivery, but only 25 percent will pay a premium for it. Thus, retailers must pay for it if they want to compete” says Treasure.

“At first, we planned to focus on the B2C ride sharing market and roll out our B2B strategy only after gaining a foothold with ride sharing. However, after discussing the current problem facing brick-and-mortar retailers, we decided to pursue a parallel path and launch our B2C ride sharing and B2B strategies simultaneously,” says Treasure.

He says they are currently in talks with several retailers looking to be competitive in a market where consumers want and expect near-immediate delivery.

The Lymousine team is currently operating as a largely self-funded entity. They received $480,000 worth of Azure services from Microsoft and will be looking to raise $3 million in venture funding.

This article originally appeared on Hypepotamus. It was reposted here with the permission of the author and editor.