I’ve heard rumors about this for a while, including from startups that wouldn’t mind one day being acquired by Amazon.
And it makes all the sense in the world.
But I’m not sure I believe it.
The main reason it’s such an open secret that Apple is working on self-driving cars is that self-driving car experts keep joining Apple to work on some secret project.
I haven’t heard anything like that about Amazon.
So if prominent autonomous vehicle engineers aren’t building Amazon’s self-driving vehicle program, then who is?
Porter’s Five Forces is one of the classic frameworks that business schools teach for evaluating the attractiveness of a business or industry.. Michael Porter, a professor at Harvard Business School, invented it.
The framework is, at its core, a five-item checklist:
Barriers to Entry
Substitution
Supplier Power
Customer Power
Existing Competition
I’m re-wording Porter’s original definitions to the phrases I think of when I step through the checklist, but that’s the gist of it.
I thought of this checklist during a recent conversation about strategy for Waymo (Google’s Self-Driving Car division).
One school of thought is that Google should position itself as the supplier of software to the automotive industry. That’s a position that’s worked well for Android, and it plays to Google’s strengths as a software company.
An alternative school of thought is that Google should build out its own transportation-as-a-service business, because car manufacturers are too smart to get trapped with Google’s software. They’ve seen how that played out for the mobile handset manufacturers, and they don’t want to get “Samsungized”.
Rather than choose between these two approaches, I want to focus here on the narrow question of whether the car manufacturers are “too smart” to get trapped with Google’s software.
The people I worked with at Ford are very smart, and I imagine the same is true at Toyota and GM and most other manufacturers.
But this is where Porter’s five forces comes in.
Some of those manufacturers are going to be way behind the curve in terms of developing autonomous vehicle software. And those manufacturers will, rationally, decide that their best bet for catching up with the pack is to partner with Waymo.
So regardless of how smart the car manufacturers are, there will be manufacturers out there who will be interested in using Waymo’s software.
Last night Udacity announced a Deep Learning Nanodegree Foundation Program, in partnership with Siraj Rival, who has been teaching deep learning on YouTube to a huge audience.
We’re really excited about this program, which is a little bit of an experiment for Udacity.
If you’re interested in joining the Udacity Self-Driving Car program, you might want to consider the new Deep Learning Nanodegree Foundation Program as a warm-up.
Any student who completes the Deep Learning program is guaranteed admission into the Self-Driving Car program, along with a $100 credit.
This year was the first CES appearance for Waymo, although it has appeared in previous years under the name, “Google Self-Driving Car Program”.
Waymo CEO John Krafcik used the opportunity to shed some light on a perpetual parlor game for self-driving car enthusiasts: “what is Google’s strategy for self-driving cars?”
Google has had self-driving cars scooting around Mountain View for years, but it’s always been unclear what the go-to-market strategy would be, and when it would start rolling.
Krafcik specifically highlighted Waymo’s gains with its in-house LIDAR sensor technology. Along with adding short- and long-range sensing units for better detection than its previous systems, the company has managed to slash the cost of its LIDAR by more than 90 percent in two years from about $75,000 per vehicle.
…
Krafcik’s language is telling — those “firsts” imply that this is just the start of Waymo’s autonomous partnerships, with many more to come. By focusing on the production of its own hardware along with software, Waymo won’t become a major automaker, at least not yet. Instead, it clearly intends to be the go-to supplier of autonomous driving systems for automakers that don’t want the expense of developing their own.
I’m not convinced the mystery has been solved yet, but it’s an important clue in the Google guessing game.
Nissan is considering building out its self-driving cars in conjunction with call centers staffed by representatives who can help the self-driving cars get out of sticky situations.
According to Wired, Nissan calls this system “teleoperation” and views it as unavoidable, at least in the short term. Weird things happen on the road and the car won’t be able to figure it all out on its own. If the car also doesn’t have a steering wheel, that leaves tele-drivers as the next best option.
On the one hand, this makes perfect sense. Wired draws an analogy to an elevator, which almost always has an emergency call button that presumably (I’ve never tried) dials somebody who can help.
The rub seems to come down to whether these call centers are staffed by representatives who only step in at critical junctures, or whether the vehicles are really “teleoperated”.
The latter doesn’t seem safe (latency being a big problem) and doesn’t seem like a big improvement over normal human driving, but it’s an interesting minimum viable product.
Drive Pilot is to the steering wheel what adaptive cruise is to stop and go pedals. Like Tesla’s Autopilot, the Mercedes system allows the driver to hand over direct control of steering and speed, while still supervising the overall operation of the car. Think of the driver as a manager in charge of employees: they’re controlling overall direction, but not micromanaging each individual operation.
And:
The system adapts to how much steering force is used, which allows the driver to decide exactly how much input to give. Use a light touch and the steering assist does most of the work. Apply a firmer hand and the system seamlessly gives up control. With Tesla’s Autopilot, applying steering force results in a slightly alarming jerk of the wheel when the system disengages. Mercedes engineers told me they wanted anyone to be able to take control of the car without any difficulty, noting more than once that the driver was always in charge, no matter how much work the car was doing on their behalf.
This is particularly exciting for me because Mercedes-Benz has been a huge supporter of the Udacity Self-Driving Car Engineer Nanodegree Program. In fact, Mercedes-Benz engineers are personally designing and teaching large parts of our upcoming Sensor Fusion and Localization modules.
Full Windshield Heads-Up Display: Oasis utilizes the full windshield to project navigation prompts and other information to the driver, while also simultaneously projecting entertainment or information to the passenger.
Autonomous Drive Readiness Check — Handover to Manual: One of the most critical concerns of autonomous vehicles is how to ensure the transition between autonomous mode and manual mode is handled seamlessly. HARMAN’s solution combines haptic feedback, eye gaze tracking, and the driver’s cognitive load readiness through pupil monitoring, to ensure that the driver is truly engaged and able to safely take control of the steering wheel.
Augmented Reality Concierge: This solution addresses the need to support increased productivity in the car while minimizing distraction. A voice-controlled virtual assistant functions as a concierge, automatically suggesting and displaying personalized points of interest while enabling advanced in-vehicle productivity to join conference calls, update calendars and more. Through Skype connectivity, the system can even translate telephone conversations — in real time — with colleagues speaking different languages.
Predictive Collision Prevention: V2X (vehicle-to-vehicle and vehicle-to-infrastructure) technologies detect objects on a collision course and offer corrective action.
Intelligent E-Mirrors: Mirrors that are automatically activated/dimmed based on user gaze.
Data is the key to deep learning, and machine learning generally.
In fact, Stanford professor and machine learning guru (and Coursera founder, and Baidu Chief Scientist, and…) Andrew Ng says that it’s not the engineer with the best machine learning model that wins, rather it’s whoever has the most data.
One way to get a lot of data is to painstakingly collect a lot of it. All else equal, this is the best way to compile a huge machine learning dataset.
But all else is rarely equal, and compiling a big dataset is often prohibitively expensive.
Enter data augmentation.
The idea behind data augmentation (or image augmentation, when the data consists of images) is that an engineer can start with a relatively small data set, make lots of copies, and then perform interesting transformations on those copies. The end result will be a really large dataset.
1. Augmentation:  A. Brightness Augmentation  B. Perspective Augmentation  C. Horizontal and Vertical Augmentation  D. Shadow Augmentation  E. Flipping
