Brad Ideas

Robocars & More

A small mystery from Robocar history was resolved recently, and revealed at the DARPA grand challenge reunion at CMU.

The story is detailed here at IEEE spectrum and I won't repeat it all, but a brief summary goes like this.

In the 2nd grand challenge, CMU's Highlander was a favourite and doing very well. Mid-race it started losing engine power and it stalled for long enough that Stanford's Stanley beat it by 11 minutes.

Uber and Volvo announced an agreement where Uber will buy, in time, up to 24,000 specially built Volvo XC90s which will run Uber's self-driving software and, presumably, offer rides to Uber customers. While the rides are some time away, people have made note of this for several reasons.

A couple of non-text interviews this week.

Robocar news is fast and furious these days. I certainly don't cover it all, but will point to stories that have some significance. Plus, to tease you, here's a clip from my 4K video of the new Apple car that you'll find at the end of this post.

On Nov 3 I flew on SAS from Copenhagen to San Francisco. The 11 hour flight had no food or drinks (other than water) due to a strike by the caterers, Gate Gourmet. It was actually surprisingly pleasant! (Unlike my experiences on Air Baltic which I will relate at the end of the article.)

SAS certainly could have done better. Since I checked in online I did not learn of this until, while waiting in the lounge, they announced that flights overseas would have "limited food choices." I figured that was not that big a deal, especially since I was seated in business class. On the way out I asked about it and they said that "limited" actually meant "absolutely none" and I was given a coupon, which I took to the airport 7-11 to get some sandwiches and snacks. There was a giant line at the 7-11 of course, but I made it to the flight.

(Trigger warning: If you only fly in coach, it will seem pompous to read complaints about problems in business class. The harsh reality is that if you travel frequently for your work, spending all that time in the compressed horror of coach simply isn't an option, especially for a big guy like me. You won't be able to do your job. So I pay the extreme prices of business class because otherwise I would not travel. If you pay that much, you expect a higher quality of flight.)

While annoying, there were some fairly positive results of the experience. The flight was actually much more pleasant without all the constant distraction of food and drink service. It makes me give serious thought to the virtue of flights, even long ones, which offer boxed food and drink that you grab at the gate and take to your seat to consume when you like.

In a major milestone for robocars, Waymo has announced they will deploy in Phoenix with no human safety drivers behind the wheel. Until now, almost all robocars out there have only gone out on public streets with a trained human driver behind the wheel, ready to take over at any sign of trouble.

Once robocars got public attention, a certain faction promoted the view that we should be giving much more attention to the idea of the "connected car." The connected car was coming sooner, would have a big effect, and some said that it was silly to talk about robocars at all without first thinking of them as connected cars. Many even pushed for the vocabulary around robocars to always include connectivity, pushing names like "connected autonomous vehicle" as a primary term for the technology.

Robocars will be connected, but not nearly as much as people in the "connected car" world imagine. And the connection won't be essential. Some cars will work with only a connection when they are parked, or with intermittent connectivity during the day. But most of all, they won't connect out to the world. The robocar probably will connect only to servers at its HQ -- the company that made it or which runs the fleet it's in. It won't talk directly to infrastructure and other cars, it may not even talk two-way with the rider's phone.

Fortunately, the efforts to require vehicle-to-vehicle and vehicle-to-infrastructure connectivity in cars are rumoured to have suffered a setback in the USA.

Rumours are swirling that the US Federal government will drop the proposed mandate that all new cars include a DSRC radio to do vehicle to vehicle communications. Regular readers will know that I have been quite critical of this mandate and submitted commentary on it. Whether they listened to my commentary, or this is just a Trump administration deregulation, it's the right step.

A few years ago, Eran Shir (who was one of my students at Singularity University and who today has an interesting startup using mobile phones to solve ADAS and self driving problems) suggested that rather than delivery robots, the future might see roving stores. These would be self-driving trucks filled with the most popular items for their region which come to you. You would open them, shop, and automatically be charged for items. From time to time they would travel to a depot for restocking.

This blog, and many other sites, paint a very positive picture of the robocar future. And it is positive, but far from perfect. One problem I worry about in the short term is the way robocars are going to make traffic worse before they get a chance to make it better.

The goal of all robocars is to make car travel more pleasant and convenient, and eventually cheaper. You can't make something better and cheaper without increasing demand for it, and that means more traffic.

I've been running this site on an ancient version of Drupal and decided to start the work of migration to something newer. Drupal 8 migration didn't work, but I have got 7 up and running, and will eventually get it to 8. Hopefully you won't see too much change other than the layout and styles. Leave notes in the comments on problems you find, but at least today I know there are a few things missing, like the topic menu and the recent comments list. I prefer a right sidebar but for now this theme is more set for a left one. The site slogan is missing too.

Nobody wants to be the first person to do or say a risky thing. One recent example of this is the revelations that a number of powerful figures, like Harvey Weinstein, Roger Ailes, Bill O'Reilly and Bill Cosby, had a long pattern of sexual harassment and even assault, and many people were aware of it, but nobody came forward until much later.

People finally come forward when one brave person goes public, and then another, and finally people see they are not alone. They might be believed, and action might be done.

Eleven years ago, I proposed a system to test radical ideas, primarily aimed at voting in bodies like congress. The idea was to create a voting system where people could cast encrypted votes, with the voter's identity unrevealed. Once a majority of yes votes were cast, however, the fragments of the decoding key would assemble and the votes and the voter identities could be decoded.

This would allow, for example, a vote on issues where a majority of the members support something but few are willing to admit it. Once the total hit the majority, it would become a passed bill, with no fear in voting.

I still would like to see that happen, but I wonder if the approach could have more application. The cryptographic approach is doable when you have a fixed group of members voting who can even meet physically. It's much harder when you want to collect "votes" from the whole world.

You can easily build the system, though, if you have a well trusted agency. It must be extremely trusted, and even protected from court orders telling it to hand over its data. Let's discuss the logistics below, but first give a description of how it would work.

Say somebody wants to make an allegation, such as "I was raped by Bill Cosby" or "The Mayor insisted I pay a bribe" or "This bank cheated me." They would enter that allegation as some form of sworn legal statement, but additional details and their identity would be encrypted. Along with the allegation would be instructions, "Reveal my allegation once more than N people make the same allegation (at threshold N or less.)"

In effect, it would make saying "#metoo" have power, and even legal force. It also tries to balance the following important principles, which are very difficult to balance otherwise:

1. Those wronged by the powerful must be able to get justice
2. People are presumed innocent
3. The accused have a right to confront the evidence against them and their accusers

How well this work would depend on various forms of how public the information is:

• A cryptographic system would require less (or no) trusting individual entities or governments, but would make public the number of allegations entered. It would be incorruptible if designed well.
• An agency system which publishes allegation counts and actual allegations when the threshold is reached.
• An agency system which keeps allegation counts private until the threshold is reached.
• An agency system which keeps everything private, and when the threshold is reached discloses the allegation only to authorities (police, boards of directors).

There are trade-offs as can be shown above. If allegations are public, that can tell other victims they are not alone. However, it can also be a tool in gaming the system.

The allegation must be binding, in that there will be consequences for making a false allegation once the allegations are disclosed, especially if the number of existing allegations is public. We do not want to create a power to make false anonymous allegations. If it were public that "3 people allege rape by person X" that would still create a lot of public shame and questions for X, which is fine if the allegations are true, but terrible if they are not. If X is not a rapist, for example, and the threshold is high, it will never be reached, and those making the allegations would know that. Our system of justice is based important principles of presumption of innocence, and a right to confront your accusers and the evidence against you.

In puzzling over solutions to gerrymandering, I remain stymied by the following problems:

• The people in power don't want to undo the gerrymandering that is putting them in power.
• The courts want to stop gerrymandering but they only overturn rules, they try not to write new ones.
• States could act, especially a few in concert, but what they do might be overturned by the courts

Arguments were heard this week in a lawsuit attempting to get the supreme court to stop gerrymandering. Courts have been ready to declare that a district is gerrymandered, but are reluctant to force states to adopt some rule on how they draw their districts. The concern is that there is no one best and most fair way to draw districts. You can tell when a district is unfair, but it's up to states to write the rules and the courts to rule on their constitutionality. The plaintiffs have a hope that the courts might rule that any districting is a disenfranchisement, and force states to allocate representatives based on a proportional system from a statewide popular vote.

I have wondered if states can find, on their own, the power to fix this. Many states have already written anti-gerrymandering rules for their own districts, but can they make those have extra-territorial effect.

One way is with an interstate compact. I outline a plan for this here. It has the problem that the counter-gerrymandering might be found illegal by the court. Remember the court is ready to say, "That's not a valid district shape" without saying a formula for what a valid district shape is. They just know it when they see it.

Here's another idea -- possibly unconstitutional as well, but it may be improved with refinement.

Recently Madrona Ventures, in partnership with Craig Mundie (former Microsoft CTO) released a white paper proposing an autonomous vehicle corridor between Seattle and Vancouver on I-5 and BC Highway 99. While there are some useful ideas in it, the basic concept contains some misconceptions about both traffic management, infrastructure planning, and robocars.

Carpool lanes are hard

The proposal starts with a call for allowing robocars in the carpool lanes, and then moving to having a robocar only lane. Eventually it moves to more lanes being robocar only, and finally the whole highway. Generally I have (mostly) avoided too much talk of the all-robocar road because there are so many barriers to this that it remains very far in the future. This proposal wants to make it happen sooner, which is not necessarily bad, but it sure is difficult.

Carpool lanes are poorly understood, even by some transportation planners. For optimum traffic flow, you want to keep every lane at near capacity, but not over it. If you have a carpool lane at half-capacity, you have a serious waste of resources, because the vast majority (around 90%) of the carpools are "natural carpools" that would exist regardless of the lane perk. They are things like couples or parents with children. A half-empty carpool lane makes traffic worse for everybody but the carpoolers, for whom the trip does improve.

That's why carpool lanes will often let in electric cars, and why "high occupancy toll" lanes let in solo drivers willing to pay a price. In particular with the HOT lane, you can set the price so you get just enough cars in the carpool lane to make it efficient, but no more.

(It is not, of course, this simple, as sometimes carpool lanes jam up because people are scared of driving next to slow moving regular lanes, and merging is problematic. Putting a barrier in helps sometimes but can also hurt. An all-robocar lane would avoid these problems, and that is a big plus.)

Letting robocars into the carpool lane can be a good idea, if you have room. If you have to push electric cars out, that may not be the best public goal, but it is a decision a highway authority could make. (If the robocars are electric, which many will be, it's OK.)

The transition, however, from "robocars allowed" to "robocars only" for the lane is very difficult. Because you do indeed have a decent number of carpools (even if only 10% are induced) you have to kick them out at some point to grow robocar capacity. You can't have a switch day without causing more traffic congestion for some time after it. If you are willing to build a whole new lane (as is normal for carpool creation) you can do it, but only by wasting a lot of the new lane at first.

Robocar packing

Many are attracted to the idea that robocars can follow more closely behind another vehicle if they have faster reaction times. They also have the dream that the cars will be talking to one another, so they can form platoons that follow even more closely.) The inter car communication (V2V) creates too much computer security risk to be likely, though some still dream of a magic solution which will make it safe to have 1500kg robots exchanging complex messages with every car they randomly encounter on the road. Slightly closer following is still possible without it.

General Motors announced this week that they would "take full responsibility" if a crash takes place during an autonomous driving trip. This follows a pledge to do the same made some time ago by Daimler, Google and Volvo and possibly others.

What's interesting is that they don't add the caveat "if the system is at fault." Of course, if the system is not at fault, they can get payment from the other driver, and so it's still OK to tell the passenger or owner that GM takes responsibility.

The Eclipse of 2017 caused dire traffic warnings, even from myself. Since a total eclipse is the most amazing thing you will see, and one was coming to a rich country where almost everybody owns a car, and hundreds of millions live within a day's drive -- I wondered how we would not have horrendous traffic. (You can see my main Eclipse report and gallery here or see all my Eclipse articles.)

Also look out below for a new 4K video I made from having 4 different video cameras running around the eclipse. I have started you 3 minutes in for the short-attention-span world, but you might also enjoy the 3 minutes leading up as the excitement builds. Even on an HD display, be sure to click through to Youtube to watch it full screen.

As described, the 4 cameras are two 4K cell phones facing forward and back, plus an HD video from a 1200mm superzoom camera and snippets of 4K video and stills from the main telescope and Sony A7rII.

The big places for predicted bad traffic were central Oregon, because it was the place with the best weather that was closest to everybody from Seattle to Los Angeles, and areas of South Carolina which were closest for the whole eastern seaboard. At a popular Eclipse site, they had a detailed analysis of potential traffic but in many cases, it was quite wrong.

The central Oregon spine around the tiny town of Madras did get really bad traffic, as in reports of 4 to 6 hours to get out. That was not unexpected, since the area does not have very many roads, and is close to Washington and relatively close to California. At the same time, a lot of traffic diverted to the Salem area, which got a nice clear sky forecast. It has an interstate and many other roads. Planning ahead, Madras was the best choice because the weather is much more unpredictable west of the Cascades. But once the forecast became clear, many people from Seattle, Portland and California should have shifted to the more populated areas with the larger roads.

I decided, since it was only 2 hours more driving to Weiser (on the Oregon/Idaho border) but much less traffic, to go to the Snake River valley. It was the right choice -- there was almost no traffic leaving Weiser. In fact, Weiser did not get overwhelmed with people as had been expected, disappointing the businesses. Many thought that a large fraction of Boise would have tried to get up to that area, but they didn't. We actually wandered a bit and ended up over the river in a school field in Annex, Oregon.

There was no problem finding space, even for free.

This is a pattern we've seen many times now -- dire predictions of terrible traffic, then almost nothing. It turns out the predictions work too well. The famous Carmageddon in Los Angeles never materialized -- even with a major link cut, traffic was lighter than normal.

This is, in turn a tragedy. It seems a lot of people did not go see the eclipse because they were scared of bad traffic. What a great shame.

4K Video

At my sight I had 4 cameras recording video. I set up two cell phones, both able to do 4K, looking at our group from in front and behind. The one behind I put in portrait mode, almost capturing the sun, to show that view, while the one in front showed us looking at the eclipse and also the shadow approaching on the hills.

Newspapers reported a house sold in my town this week for almost $800,000 over asking, which is to say the buyers bid $2.5M for a house listed at $1.69M. Now the prices are already crazy but this takes the cake. The buyers had lost a few auctions before by not overbidding enough, and wanted to make sure they got this house, even though it is not that remarkable a place.

Our videoconferencing tools have been getting better, but meetings with remote video participants still don't work very well. One problem is poor use of the technology (such as a lack of headsets) which I outlined in my guide to room based video meetings. These can be worked on and the tech keeps improving.

The other big area for improvement is the discipline of the people in the meeting. The big challenge in typical meetings is that some of the participants are 2nd class. This is obvious when you have a meeting room with multiple local people and some remote users. It can also happen when people have differing levels of technology. In an ideal meeting, everybody in the meeting is on the same footing as far as their presence and ability to communicate. In addition, everybody should be as fully engaged with the meeting as if they were in a single olde-tyme meeting room.

We break this rule often. It is quite common to have remote attendees turn off sending video, or mute their audio, for example, making them be more like a TV audience than members of the meeting. It makes sense because it saves bandwidth, and people don't like being watched. We also tolerate having some people present just on the phone, while others are there in person and others are on low and high quality video systems.

If you hope for a good meeting, you also want to express that the main value of the conferencing system is to let people attend without travel. It is not there to let them attend without the same effort and engagement they would put into a meeting they did travel to. The things I describe may seem minor, and they may veto features of great convenience, but those features are actually bugs and disrupt meetings more than people realize.

Here are some principles to get around this:

No meeting room

In an ideal video meeting, everybody is on their own personal video station. There is no meeting room. This means that even if several of the attendees are in the same building, they don't go to a room, they stay at their desks and join the meeting just like any other remote.

This is obviously hard to do if the majority of participants are in the building, but it can be worth it. It also means you don't need room-based videoconferencing systems, which are expensive and don't work well. But if only 2 or 3 of the participants are in the same place, definitely consider having no meeting room. The big benefit is that when everybody has their own microphone, everybody hears everybody really well.

Today you can't have people in the same room using their own computer because they hear the other people both via their headset and through the air. Perhaps some day a smart videoconferencing system will understand that some people are in the same room (you can tell because some sounds do get into the microphones) and adjust. It would allow those who still want a physical meeting room to get the great audio and video that comes from everybody using their own computer and headset. Those in the room together would still be higher-level participants, but remotes would not be that badly off.

Headsets at all times

We have gotten seduced by how well some voip systems handle speakerphone mode in one on one conversations. Don't be fooled. They don't do group meetings well at all. They seem like they do, but quickly you realize that now everybody hears all the random noises from the location of a speakerphone user. They do things like step away from their desks to eat, chat or take a phone call, and everybody hears it. Keyboards and mice clickety-clack. Sirens go by. It's easy to ignore this in a one on one call, but it disrupts a meeting.

A new organization named Equal Votes is pushing to make a supreme court case to undo the electoral college. They hope to use a precedent set in the famous "Bush v. Gore" 2000 election case, which strengthened the application of the equal protection clause to election law. They want to show that the "Winner takes all" approach that 48 states use to hand out electoral college votes is a violation of the idea of one person one vote. States would then not be able to use it.

It's an interesting idea, but I have grave concerns that it might backfire, and badly. Worst case, it could guarantee a Republican President into the indefinite future. Hopefully not, but that's a bad result, even if you're Republican leaning.

My friend Larry Lessig is supporting this, and I don't doubt the theory that it might win. The idea is that since a state could assign its electors in proportion to how its citizens voted, it is depriving them of their rights by not doing so. If you are in a safe state and not in the dominant party, your vote is useless and effectively counts for nothing.

It is far from certain that this logic can win. Winner takes all is of course the norm in democracies. The country only gets one President even though almost half the country didn't want him. States only get one governor. In California, where state legislators are elected by districts and some Republicans get seats, the reality is that the Democrats get complete power, not partial power. And in the current congress, it takes defections or filibusters to stop complete rule by the Republicans (even though, due to gerrymandering, they got fewer votes than the Democrats in the house.)

Voting by gerrymandered district

And this is where we get the first big risk. Two states, Maine and Nebraska, don't do winner takes all. They follow what seems a very sensible plan at first. A state's electoral college delegation is equal to the number of congressional representatives they have: House members plus two. These states select two electors based on statewide vote, and then select the other electors based on the vote in each district. Just like members of congress, the electors represent their "districts."

There is a risk that if the court held that winner-takes-all is invalid, they might not rule that the established district based system is invalid. Equal Votes will ask for that, but they might not get it. If they don't get it, then some states, in particular the gerrymandered states, may decide to follow the Maine system. And being gerrymandered like this, they are likely to return a slate of electors that is not too different from their congressional delegation -- a heavily skewed delegation. Even if many other states use a different system (like the proportional allocation Equal Votes wants) the result would be a college very much like congress. And that's a GOP college.

Why? Because of operation Redmap. This insidious, I would say evil, cheat on the electoral rules was done by Karl Rove and the RNC in 2010. They poured millions into a carefully selected number of unsafe Democrat statehouse seats around the country, enough to swing those statehouses to the GOP. Then they used that power to redraw the districts in those states in a gerrymandered way to favour the GOP. Not just in the congressional elections, but in future statehouse elections, cementing the power indefinitely. It will need a major anti-GOP swing, or supreme court ruling, to fix it.

The robocar revolution has the potential to assist China in dominating vehicle manufacturing. That's the bad news -- unless you are a Chinese manufacturer. The better news is that manufacturing is only part of the car industry, and it's getting smaller.

• China has the largest car manufacturing industry, and is strong in electric cars
• Brand of the manufacturer is almost irrelevant in taxi service
• Reliability of the taxi is much less relevant
• US tech companies need manufacturing partners
• The money in ground transport is in service, not cars

Today, Chinese brands are not sold in any numbers in the USA, or almost anywhere outside of China, but China is already the largest car manufacturing country in the world. Chinese brands have no cachet (even in China, it seems) and western and Korean/Japanese brands are strong. How might that change?

Car brand is very important for people buying a car to own. In fact, the nameplate is the top source of value in a modern car sale. The difference is that we will be moving from people buying cars to own towards people buying rides.

When you order "Uber Select" (Uber's nicer-car offering) you don't care if what shows up is a Lexus, BMW or Mercedes. You don't even car if its a Hyundai Genesis, their brand-new attempt at making a luxury marquee. You are only going to ride in it for 15 minutes. It has to be comfortable, smooth and look nice, but rarely does the logo on the outside matter.

It's the Uber brand that matters (though not as much, as most people would find no difference between an UberSelect and a Lyft Premiere as far as the vehicles are concerned. And you might not even care if it's a Great Wall Wey (a Chinese luxury car you've never heard of) that picks you up if it looks nice and gives a reliable ride.

Of course, today the top makers like Mercedes, BMW, Lexus, Audi, Acura, Infiniti and others are known not just for luxury, but for quality. They make well engineered, reliable cars in a way the Chinese are not quite ready to do.

But do they have to? If your expensive BMW breaks down, you have to get it towed, arrange its repair and get a rental car. You're pretty angry at BMW when it does, and you paid a lot for that car to avoid that experience, and usually you do. If a car in a robotaxi fleet breaks down, you're very unlikely to even know it happened. Very rarely, a car like that might break down when you are riding in it. It would pull to the side of the road and have already summoned a replacement car. Within 2-3 minutes a new vehicle will pull up and take you on your way while the company sends a tow truck to deal with the broken car.

Of course, if it broke down while on its way to you, might might not even know it. But even the breakdown while driving will be barely worth mentioning to friends, it just didn't inconvenience you very much at all.

While the BMW will surely break down less than the Great Wall (at least for now) it also costs a great deal more. That might be worth it to avoid that owner's breakdown scenario, but it's not for a fleet breakdown. For a fleet manager, it's just a question of whether vehicle downtime cost is more or less than the extra cost of more robust engineering, with a small factor for customer inconvenience.

To top things off, I predict robocars will have fewer breakdowns. They will always been monitoring themselves, and will come loaded with sensors. They will always get proper maintenance, taking themselves to maintenance depots when it is needed. They will test all systems like brakes, steering, tires, engines and more every day or every hour when running vacant. They will never let anything get too hot or vibrate too much. Both the BMW and the cheap car will do that.