A Case For Tesla’s Ride-Hailing Service – A Pre-Robotaxi Option
Seeking Alpha Analyst Since 2020
Summary
- When Tesla starts producing quantities of their new $25k compact EV they should gradually introduce a ride-hailing service city by city to ensure a highly level of service.
- Tesla needs to collect real-world taxi driving data which is different and a higher quality data than the driving data of Tesla owners because it is directly related to Robotaxi.
- The frequency of interventions data by the driver in ride-hailing service in complex city traffic environments is the type of data that most likely will be required by regulators.
- Apartfrom physically driving the vehicle, taxi, Uber and Lyft drivers make hundredsof decisions daily related to being a taxi service that owners of EVs seldomencounter. Robotaxis need to learn how to make these decisions too.
- As Tesla’s AI masters computer vision for autonomous driving,they will also have to master voice recognition and natural language AI toprovide the comparable service or experience to its’ passengers in a Robotaxiservice to be the same as or better than a taxi or ride-hailing service.
Tesla’s Ride-Hailing Service - Ark Invest has been advocating that Tesla launch a ride-hailing service like Uber or Lyft before their Full Self Driving (FSD) software is ready for the launch of a Robotaxi autonomous fleet service. I agree that once Tesla begins mass producing the $25k compact EV, they should launch a ride-hailing service gradually - city by city, perhaps starting in San Francisco. I expect that 75% to 80% of the Tesla Robtaxis will be owned and operated by Tesla and 75% to 80% of the Robotaxi fleet will be comprised of the $25k compact EV because the average car trip has about 1.4 passengers and the average person drives less than 30 miles per day. Using the bigger Model 3 and Model Y EVs with bigger, heavier, and longer-range batteries would be very inefficient for most of the Robotaxi fleet.
Tesla could learn a lot about the ride-hailing business and improve their ride-hailing app, customer service, and business processes. First impressions are critical to the success of a new entry in a crowded marketplace. Tesla can ensure a high level of customer satisfaction with shorter wait times, lower costs, high availability (especially in metropolitan service areas), higher quality of service and nicer cars by hiring drivers as Tesla employees or contractors with fixed schedules. Drivers with a fixed work schedule will allow Tesla to build complex algorithms that optimize the number and placement of Robotaxis based on predictive modelling data, traffic data, and vehicle data (battery level, location, vehicle type). This will allow Tesla to provide superior service compared to the incumbents.
These drivers could drive their own Tesla EV, lease a Tesla EV, or drive a Tesla owned EV. When Tesla gets regulatory approval to launch their autonomous Robotaxi service in each city or municipality, the existing fleet of ride-hailing Tesla EVs would eliminate the Robotaxi ramp-up phase and provide a smooth transition to their driver-less service especially if the regulatory approval process takes longer than expected. Customers already familiar with Tesla’s ride-handling service are likely to be early adopters of the autonomous service. Tesla could retain some of the ride-hailing drivers for their customers that might initially be a little skittish about a driverless Robotaxi.
Tesla could reward the layoff drivers by offering them deep discounts on the purchase of their Tesla EVs, which the drivers could enlist in the Tesla Robotaxi network and enjoy passive income for the life of the vehicle.
While Tesla waits for the perhaps a slow and laborious municipality-by-municipality regulatory approval process, Tesla can expand their ride-hailing service to other cities across the country, so that by the time they receive regulatory approval, all of the Robotaxi vehicles and the charging/cleaning stations infrastructure will already be in place and ready to operate autonomously with the flip of a switch.
More Data and More Quality Data! – The ride-hailing business would allow Tesla to collect high quality real world driving data especially in the busy and complicated city traffic environments. Many people think driving is driving is driving. Once you know how to drive on any street or highway, the rest is easy.
A taxi, Uber, or Lyft driver has unique driving experiences not experienced by most Tesla owners that drive their cars from home to work in the city every day. Tesla will learn the nuances of a taxi service by gathering real-world driving data from a ride hailing service in order to train their Robotaxi fleet. The number of interventions that a ride-hailing driver will experience daily will be much greater than the interventions experienced with a Tesla owner commuting to work. Regulators are much more likely to approve an autonomous taxi service if a large portion of the data consists of actual taxi driver or ride-hailing driver miles.
Taxi Service Operations Data – A Tesla ride-hailing business will also provide a great deal of the taxi service business data that will help Tesla develop a better service for its’ customers. Taxi drivers and ride-hailing drivers make hundreds of complex decisions daily that are unrelated to the physical act of driving the car. Most of us take these actions for granted but would cause big problems for a computer driven vehicle. Below are only a few examples of the decisions taxi drivers and ride-hailing drivers make daily:
- How long do you double park (blocking traffic) to pick up your next passenger or do you waist 10 to 15 minutes circling around the block because the passenger is not there yet.?
- When dropping off an elderly couple at the airport with several heavy bags of luggage, do you pull over at the first available spot, 50 or more yards from the entrance, or do you wait because you see a car much closer that looks like it’s about to leave? How long do you wait? Do you double or triple park illegally for a few minutes (in a No Parking, No Waiting zone)? What do you do if a car that you are blocking wants to leave and the elderly couple is taking a long time to unload?
- What do you do if the passenger looks drunk and out of control? Do you pick him/her up?
- What do you do when a passenger suddenly goes berserk in the car? Do you drop them off immediately? Do you drive them to a safe place and drop them off? Do you take them to the Police Station or Hospital?
- What do you do if a passenger leaves a bunch of trash, vomits, or quietly urinates or defecates in the Robotaxi? Are there enough sensors and cameras to detect this?
Tesla will not have this data without a ride-hailing service pre-launch.
The SERVICE Business – Tesla has very little direct day-to-day customer service experience. Tesla customers can buy a Tesla in less than a minute using the Tesla App. They can have their car delivered to their home or office in a touchless (COVID-19 safe) exchange. Tesla’s App and their in-car training videos can instruct the customer how to operate the vehicle. Through all of this, Tesla provides one of the most pleasant car-buying experiences for its’ customers (compared to going to a dealership spending hour haggling and filling out paperwork, etc.). Unfortunately, Tesla has very little experience providing constant minute by minute, face-to-face customer service with a public that is at times fickle, illogical, spontaneous, erratic, emotional, irrational, indecisive, and stupid. It doesn’t matter if a Tesla Robotaxi is cheaper, faster and much safter than any other transportation alternative. If Tesla’s Robotaxi drop-offs an elderly couple with heavy luggage, 50 yards from the entrance to the airport, or if the Robotaxi takes 15 minutes to circle around the block in city traffic just because it is illegal to double park, and makes the passenger late for an important meeting, Tesla’s Robotaxi service will fail. First impressions are so important to a new entry in a crowded marketplace. Videos of a few bad experiences with new fledgling technology or service goes viral much faster than thousands of good experiences.
Voice Recognition – Of all the videos that I have watched on Tesla’s FSD limited Beta release, none of the drivers are beta testing the voice recognition system (probably because it is not available yet). A robust voice recognition and response system is critical to the launch of a superior Tesla’s Robotaxi service. Tesla will learn how often a passenger needs to talk to the driver to change the destination, to make stops along the way, or to respond in real-time to events as they happen. Passengers are unlikely to want to restrict all of their communications and interaction with the Robotaxi through the Tesla App especially in real-time driving situations. How many times does somebody hail a taxi and isn’t sure where they want to go. They may be in a rush but need time to think of where they should go first. The following are a few examples of voice communications a Robotaxi probably will need to respond to:
- “Slow down…”, “Is that my wife/son/daughter/boss/co-worker/old friend.”
- “Stop here!”
- “Wait a minute!”
- “Turn here!”
- “Turn around.”
- “I changed my mind, take me home.”
- “Taxi!, Taxi!, Taxi!”
- “Hurry-up, I’m late for my flight!”
- “You are driving too fast!”
- “Is that a bear!”
- “Pull over!”
- “I need to pick-up my husband/wife/children/friend/co-worker…. along the way.”
- “Where can I find a good sushi restaurant near here?”
- “There must be a good happy hour near here.”
- “What is happening this weekend?”
- “Where are the best bars in this city?”
- “What’s the weather forecast for today?”
Launching a ride-hailing service will allow Tesla to test and improve their voice recognition/response natural language system by initially simulation the computers response by running in shadow-mode (the same way Tesla currently runs Autopilot simulations in real-time shadow-mode to compare the computers decisions to the human driver operating the vehicle). As the natural language AI gets better, the computer should be able to handle conversations with the passengers and drivers could intervene whenever the Robotaxi has trouble understanding or responding. In addition to voice recognition, Robotaxis need to respond to beeping cars, police, ambulance, and fire engine sirens. It’s better to work out all of the kinks in a ride-hailing environment than work them out after the launch of the Robotaxi service.
Taxicab Therapist – The holy grail of a Robotaxi service is the Taxicab Therapist. Ask any bartender, taxi driver or barber/hair stylist and they will tell you that serving drinks, driving from one place to another or cutting and styling hair is only a part of their job. The most important part of their job is listening to, talking to, and interacting with their customers. A Robotaxi needs to sense their customer’s mood by analyzing their tone of voice, word choice, eye movements, heart rate, breathing patterns, and body language to determine the passenger’s mood and how to respond. Is the passenger happy, sad, irritated, chatty, flirty, curios, inquisitive, an introvert, an extravert, depressed, tired, self-involved, or just wants to be left alone. Humans are social animals and typically crave interaction. Some are lone wolves and would rather be left alone, but most crave human interaction. For many people, a daily 20 minute or longer commute in a quiet and sterile robot-driven car can be a sad lonely experience. Many customers would prefer to take perhaps a slower, less safe, and more expensive human-driven ride for the human conversation and interaction.
A ride-hailing service would allow the Tesla’s AI computer to learn how to read human behavior and respond accordingly by analyzing the behavior of the driver and the passenger and shadow the interchanges in real-time to improve the AI’s interactions.
As we go further down this AI rabbit hole, we can image that the Robotaxi that you take every day to work begins to know you and understand and becomes your friend – like your personal chauffer or like the bar tender at your favorite bar. The next logical step is to imagine that Tesla leverages their AI and data platform advantage to launch a Personal Assistant (Tesla BFF) App on your cell phone that you can talk to, confide in, trust, and depend on all day long (like the movie HER).
WOW!.
Mike Drop!
Analyst's Disclosure: I am/we are long TSLA, Appl.
Seeking Alpha's Disclosure: Past performance is no guarantee of future results. No recommendation or advice is being given as to whether any investment is suitable for a particular investor. Any views or opinions expressed above may not reflect those of Seeking Alpha as a whole. Seeking Alpha is not a licensed securities dealer, broker or US investment adviser or investment bank. Our analysts are third party authors that include both professional investors and individual investors who may not be licensed or certified by any institute or regulatory body.