Hacker News new | past | comments | ask | show | jobs | submit login
Nobody Wants to Let Google Win the War for Maps All Over Again (bloomberg.com)
154 points by adventured on Feb 22, 2018 | hide | past | web | favorite | 103 comments

On any given day, there could be a half dozen autonomous cars mapping the same street corner in Silicon Valley.

It seems kind of tautological that they're using autonomous vehicles to create maps because they need maps for autonomous vehicles.

Yeah, I know, specialty equipment... but still.

Actually, what seems really odd is that new maps are that big of a deal. The cars need maps for normal navigation, sure, but you wouldn't expect that they would rely on anything too precise for safety reasons alone. The cars need to be able to adapt to possibly changing conditions and can't misbehave (at all) just because a street closed, cones are on the street, etc.

Unlike conventional digital maps, self-driving maps require almost-constant updates. The slightest variation on the road—a construction zone that pops up overnight, or a bit of debris—could stop a driverless car in its tracks. “It’s the freak thing that happens that’s going to make autonomous not work,” said McNally, the analyst.

I would call an autonomous vehicle that couldn't handle some debris "fundamentally broken".

Seriously, this mapping thing smells like a boondoggle.

As someone who's old enough to remember road trips armed with nothing more than a Rand McNally road atlas for navigation (really not that long ago - and I do continue to take one along nowadays as a backup), I'll never trust an autonomous vehicle that needs to be fed a constant stream of up-to-the-minute mapping data in order to make decisions.

I am with you here. Autonomous vehicle may need GPS for navigation, but it should be able to drive down the roads without it.

Definitely. GPS can't be relied on: there are lots of places with naturally bad reception and the military sometimes tests GPS jammers, like during the recent exercise.

It's interesting to note that the first in-car navigation system used no GPS: it relied solely on inertial navigation and map-fitting to determine its location. For navigation, an autonomous car needs something like that as backup for GPS. All the actual driving needs to be done with on-board sensors, though.

Dead reckoning is actually fairly accurate if done right.

I'd be surprised if the lidar units aren't being used for a similar algorithm.

Doing dead reckoning "right" over an extended distance requires extremely expensive laser ring gyroscopes. It's going to be a long time before those (or something with similar accuracy) will be affordable for consumer use.

Would a self-driving car need perfect dead reckoning? I was able to update my assumptions about my distance and location based on street signs before I had GPS...

Couldn't the device correct using local land marks? A minimal offline map, maybe even just topographic, that shouldn't change too often. "Expected hill was crested 5 meters sooner then expected, update estimated position accordingly".

If you have local land marks then you don't really need dead reckoning at all. Cheap dead reckoning sensors have so much error and drift right now as to be mostly useless for driving. The expensive sensors work but aren't economical for cheap passenger cars.

Here's the system I was talking about. It used dead reckoning and landmarks (turns) to periodically re-calibrate the fix. This was all done with 80s technology.


Live demo: https://www.youtube.com/watch?v=CHCCjlSWbHE

> Cheap dead reckoning sensors have so much error and drift right now as to be mostly useless for driving.

I don't think anyone's saying that it would be used for the driving part (i.e. car control), just as an option for orienting the car in the wider world beyond its direct perception.

GPS is a nice-to-have, but wouldn't be necessary for a true AI, given an autonomous car can have more and better on-board sensors than your basic-issue human (we don't have eyes in the backs of our heads, or LIDAR).

Why would a True AI (tm) need a car in the first place? "Give me your clothes your boots and your motorcycle" and all that.

I don't see how GPS is a "nice to have". How does an entity know to get from A to B without some sort of navigation assistance?

Unless your "true-AI" stops to ask locals for directions, its going to need some sort of GPS. Can't say I'm looking forward to arguing with my car about how it should have taken a left on Main St, instead of a right.

> I don't see how GPS is a "nice to have". How does an entity know to get from A to B without some sort of navigation assistance?

GPS ≠ navigation assistance. You can infer where you are on a conceptual map from reading your surroundings, especially signage. Getting a latitude/longitude estimate from the GPS network and trying to match it with an overly precise and possibly out of date map isn't the only way to infer your current position. It's only useful when you don't already know where you are, and computers are good at precisely remembering the path they took from the last known observed reference.

Plus, self-driving cars need to be aware of their surroundings, and not solely rely on blindly following a hyper-precise map as a prescription of the road details ahead. Having the local observations become the canonical precision knowledge, and keeping the map as conceptual knowledge of how roads connect places, implicitly gains a lot of this sort of benefit.

I mean rear-view and side mirrors are the eyes in the backs of our heads, though a level 5 self-driving car would be as you say.

Brains pretty much need a constant stream of up-to-the-minute data as well: eyes and ears, both of which can fail.

That's not what he's saying. He's saying a stream of data from a server rather than directly from sensors.

Ah. Well, they don't need a stream do they? Why not periodic downloads of the map area 10 miles in every direction?

Safety is not a boolean value, you have to look at this from a risk perspective.

If you have a self-driving car that's pretty safe, but you can make it even safer with maps, and those maps are going to be accurate 99.9% of the time, the 0.1% of the time will be less safe, but overall you've significantly reduced your risk of an accident, and maybe you can even mitigate the risk by doing things like slowing down when your maps aren't accurate, the same way human drivers slow down when there are cones on the street.

The issue isn't so much that they "couldn't handle" some debris, but that "handling" something is not a yes/no proposition.

Having said that, I haven't worked for these groups, so I don't really know what these maps are used for, maybe these maps are just used for localization because GPS isn't accurate enough.

I understand all that; but in terms of liability, safety is a binary issue. There will be tolerances for safety that can't be lessened because there's a tire in the road that wasn't mapped yet. "It wasn't mapped yet" won't be acceptable, so the vehicles will need to maintain a lowest-common-denominator safety profile at all times based upon their sensory input - not upon networked debris information.

The debris type information should only be useful in terms of routing.

I don't think anyone outside the autonomous groups really knows how map dependent they are.

It's easy to imagine reasons to have detailed maps in the areas they are operating even if their systems are not wholly dependent on the maps. Stuff like being able to more quickly analyze how well the system is characterizing the environment or reducing power consumption or the map is necessary above a certain speed or …

GM made a detailed map of limited access highways in North America just to geofence the super cruise control in the new Cadillacs, so I'm also not sure the duplication of effort matters all that much.

> you wouldn't expect that they would rely on anything too precise for safety reasons alone

I don't think they do. It's just more data that gets mixed in with multiple streams of real-time sensor data.

> Seriously, this mapping thing smells like a boondoggle.

It makes sense to me that autonomous cars would simultaneously be map users and map creators because I assume that data is shared with other vehicles. It's going to drive the cost of the map making way down. I think eventually the company with the biggest fleet will have the best maps, perhaps even real-time. Waze does this now but it relies on user reporting.

You need the map because you want to know where to go, not in order to avoid debris.

Pick two points in your town: say your home and your office. Imagine a stranger had to get from one to the other, and they know how to drive. The map if for all the information that stranger does not have.

Why do you need anything with more precision that current-day Google maps for that?

That is a different question about how often they should be updated, what sort of updates, when does construction happen etc.

You don't need a map for that, just a sense of direction and distance. I have often navigated someplace by making seemingly random turns - I knew where I was going was southeast so any east or south road was fine. Once I got close street signs got me the rest of the way.

> You don't need a map for that, just a sense of direction and distance.

Without map you also need time. In some cases, a lot of it. Not all cities have perfect grid structure and there are one-way streets that can mess you up a lot. In the city where I live it would be really time consuming, because there's a huge pedestrian-only area in the center and many the streets around it are one way, or turning is prohibited at some crossroads so you cannot suddenly decide that you made a mistake and you want to make a U-turn and go back.

Without maps you need time or luck.

That works in cities that are perfect grids. Try navigating the scottish highlands like that. You might have gone miles down a road before you realise that it's not going the way you want to go.

If you are from the US you may be used to use cardinzl points. I was often given directions like "the south side of the building" or "go north east" when visiting the US.

This is not common at all in Europe. Sure, we have a general idea of the direction ("it's with of Paris") but we would not use that for anything else. Especially the building case was weird to me, I woukd have said "the side chose to the entrance" rather than "SE"

> ... just a sense of direction and distance

OK, how does a first time visitor with a rental car get a sense of direction and distance without a map or driving directions from a local?

How do you reliably solve the beginner's mind problem for a self-driving computer?

Without a baseline input, the car will have to rely on the passengers to do back-seat driving which defeats the point of driverless.

Governments should open all their mapping data, as some have already done. Probably in quite many countries, there are also entities collecting more detailed data that is used for the maintenance of the roads. All this should be opened up as well. I believe just giving out the data freely will benefit these countries more than trying to extract some small profits by selling it.

"Governments should open all their mapping data"

Most don't have a lot of useful data they collect/maintain themselves. It's often easier and cheaper to buy it from some party who gets others to pay for it as well, spreading costs. I used to think like you do, back when I didn't realize just how much work it is to maintain, and then to provide it to all stakeholders in the formats they require. Here in the EU there was a big push 10-ish year ago to make all this information free, at various levels; see e.g. the INSPIRE directive and all work that that caused. Nowadays people have experience with just how much work it is, and have experienced that the promised benefits didn't quite realize (for various reasons). Most enthusiasm is gone, I feel.

(as a researcher, all this data becoming free/easy to download has been a huge quality of life improvement for me; but no more than that. I don't feel there is that much new stuff going on that wasn't possible before. Sure, before you had to chase people for weeks and sign all sorts of agreements, but the end results were more or less the same for professionals. Interested amateurs at home would give up in the past, but they didn't/don't produce all that much useful anyway.)

The US does have open mapping data:


While useful for taking the US Census, the TIGER data isn't really suitable for most navigation tasks. At best it can be used as a base layer to build a geospatial dataset.

OpenStreetMap did this about a decade ago:


That being said, the current administration is looking to cut TIGER funding, not increase it.

For the foreseeable future getting mapping data suitable for navigation in the US is going to be the job of private for-profit companies.

I've read that a lot of the limitations/errors associated with mapping data once you get away from fully maintained paved roads stems from limitations in TIGER data. You get into very secondary roads and there's not much that's really authoritative to distinguish all the gradations between unpaved but well-graded and maintained and in appropriate weather you can drive it with high-clearance 4WD if you know what you're doing.

For a navigation system those roads are mostly a "last mile" problem. They generally aren't as direct or connected as main highways, so graph based navigation naturally ignores them. Main highways are also reasonably well marked in the tiger data, so the lack of differentiation for the lower priority road type doesn't impact a system that prioritizes larger highways (a pretty reasonable heuristic).

The lack of differentiation is a big problem if you want to pick which way to go when driving around though.

I mostly agree with all that. Although in some areas like the rural US West, very secondary roads are pretty common. This becomes a bigger problem as more and more people depend on computers to tell them where to go without thinking much about it.

Of course, this isn't really a new problem. People have died following roads on paper maps that weren't suitable for, say, winter travel.

For a car navigation system, yes. For a bike navigation system it's a massive problem.

Hybrid approach: a compromise like the original public-private copyright deal, might be best. You tender your data to the government. Government publishes it but gives you exclusive rights for N years. In exchange, for the first 2N years, only those who have contributed can access others’ data. Maybe throw in some tax deductibility if the incentives aren’t strong enough.

The data that is used to compose a map is not the same data you need for navigation. Generic map data is freely available for pretty much everywhere.

In most countries, private companies’ (particularly Google) mapping data is far better than government data.

You can't trust governments to have updated and reliable data.

That depends entirely on the government and the set of laws and incentives they operate under.

I found it mildly amusing to see an article about concerns over data monopoly on a Bloomberg website.

A good example of new vs old generations of data monopolies.

But, I think we need to stop thinking of the tech monopoly problem in principled, abstract terms. That's hard because it's how law works, and how our moral instincts genrally work.

Bloomberg's data monopoly does/did pay well, give them a nice moat and such. But, it couldn't be leveraged in the same way modern data monopolies can. It didn't permeate into new areas every day. It didn't have all these consequences (eg killing privacy) of scale. They din't lock up a dominant position in categories (eg smartphones, cars) that wouldn't exist for a decade or three.

Bloomberg's market position is more conventional. Being a monopoly gives them pricing power, and nicer margins. That is tame by today's standards.

Today's monopolies can't even be reasoned about in the language of antitrust. they don't act as trusts, generally. Market share is obscure. How do you determine the effects of reduced price competition for maps? It's free. The whole logic of econometrically calculating the harm to a consumer is irelevant.

Yet, the idea that Google owns your phone, browser, search engine, drives your car and controls the ads you see 100 times a day.... That's still a worrying thoguht. Economically, it's easy to see (eg search->maps->phones->cars) how one monopoly leads to another and they strengthen eachother.

They don't have monopoly on the data, just on the access. And they start to realise that this can disappear quickly if someone invests enough to provide a competitive interface (e.g. Reuters). Google's in a better position, replicating underlying data is even more expensive.

It's worth stepping back and considering the landscape of self driving.

First, the fact that it's within reach is amazing. It's one of those wierd cases where regular people instinctively put it in the "mars colony" category of futurism while sober insiders had to bring us back to a more optimistic view.

Now, there is a lot to worry about in the current landscape of giant tech companies, and this article is another example. But, there are things to awe at too.

Having self driving within reach... it's a result of massive effort. Amazing technology, and complexes of technology it is built on. This includes the manufacturing technology producing the cheap components.

All of this cam from (1) fairly blue sky intiatives (2) with extremely long and uncertain payback periods and (3) tremendous strategic money-where-your-mouths are foresight. When I realized what G where doing with street-view, I thought it was crazy. More a sign of overexuberance and too much money. Turns out, it was en route to maps dominance which was en route to transportation dominance (and phone dominance).

IDK if this is unprecedented, but it's fairly unrivaled in today's world. We don't see IBM or Intel, Toyota or McKesson thinking this big or this far. Definitely not banks or money businesses. Not cities or countries, generally.

I mean an equivalent in health/medicine would be taking the ideas from TED talks. A personalized health revolution. Nanobots or curing death like Aubrey de Grey keeps talking about.

I'm already hearing the "regulations" explanation. Medicine is restrained by surgeon generals, parliaments and national health setups. I don't think that argument holds at this scale. Self driving cars require re-rwriting our transport rules, and probably massive the physical infrastruucture. The technology will have sucked up billions decades before it makes a penny. The guiding idea is that "if we have a true self driving car, the other stuff will sort." Same applies to medicine. If it's better enough, it'll take.

It's hard to peer back and determine of something was inevitable or not, but I suspect self driving wasn't. It could have happened (could will happen? tenses are hard) decades later, if not for the fairly outrageous decisions of a small number of people.

> First, the fact that it's within reach is amazing. It's one of those wierd cases where regular people instinctively put it in the "mars colony" category of futurism while sober insiders had to bring us back to a more optimistic view.

I think a Mars colony is closer than self driving cars. You need basically strong AI for fully self-driving cars. You just need a ton of money and an incentive for Mars colonies. Most of the science for going to Mars is already there, as well as the tech. I don't think we even truly have the science for wide spread, all-weather, all-terrain self driving cars.

Self-driving cars will be in wide use long before every case is solved. There's lots of money to be made with autonomy that only works on well-maintained roads in regions with favourable climate.

Even highway driving in clear weather conditions that allows you to truly go hands off and read a book would be a huge win. For a lot of people, it's a large amount of the driving they do and I suspect it's where a lot of the worst accidents occur because of the speeds involved.

Of course, it doesn't get you to the futurist vision of self-driving cars in that you can't share cars and you need a licensed, sober human in the vehicle but it has the advantage of being very achievable in the fairly near-term while still being a really nice feature.

Yeah, self driving cars may work in Arizona but how about rush hour Amsterdam or New York? In my country it might work for long haul trucks on highways in the near future. But I dont see a use case beyond that.

And rush hour Amsterdam or New York are probably easy mode compared to rush hour Cairo or Delhi :)

A lot of self-driving thought seems to go into deciding when to obey rigid rules and when to break them like everyone else.

We can't assume the pure chaos of Cairo or Delhi requires the same model. Self-driving might only need basic Roomba-level collision avoidance and basic pathfinding for a place where the humans navigate with the expectation that every object, moving or not, is a collision waiting to happen.

Think about how much of rules-based driving is about avoiding hitting people who aren't following the rules. If you start off assuming there are no rules, the model is simpler: everything the sensors pick up could come flying at you at any moment, so plan accordingly, and stop or slow down if you can't find a path that avoids an impact.

Hard to say what roadblocks will be there, 3 steps ahead. I suspect that if the US' easy cities become self-drivable, all cities will. But, maybe not.

In any case, I don't think we should be assuming infrastructure will remain the same. If self driving does replace driving, the infrastructure will change. Point problems (eg, computer can't handle these 43 tricky roads or intersections, those problems will be solvable be changing the road. It might just be adjusting so gns or minor parking rule changes.

There's a chicken-egg problem, which is why getting some of the way on existing infrastructure is smart. It also avoids scaring finance ministers, but ultimately infrastructure will meet driverless cars part way.

I suspect that "success" requires getting most of the way on existing infrastructure, especially to the degree that "self driving" is mostly for people who are well-off financially for a significant period.

That said, we redo intersections because they have a lot of accidents all the time. I'm not sure why intersections that self-driving vehicles have particular problems with would be any different. I can also imagine having some sort of RF "beacon" that gets used for construction sites and the like to flag drivers that they need to take over manual control in a mile or whatever.

To be fair, the chaos the traffic situation is probably inversely proportional to the labor cost of driving.

I can't help but think that self-driving cars is the penultimate American thing. They aren't effective or needed in properly built cities (think London) and they won't make THAT much good to the society eliminating taxi drivers. If we dig into the vision of what people want to achieve with them, it's mostly manifestations of American problems (poverty, systemic racism, homelessness, segregation, insular indifference to others' problems, crazy urban sprawl) which people don't want to confront directly, instead opting for "let's spend HUMONGOUS amount of money to keep everything more or less the same".

There is a solution for most problems people try to solve with self-driving cars: more cycling, better public transport, denser cities. Look at London: car use is heavily discouraged (20mph speed limit, congestion charge, expensive parking) and still the whole thing works like a charm.

You're missing many advantages of self-driving.

1. Better safety - I find it very likely that self-driving cars will reduce the amount of traffic incidents. Algorithm will likely act more stable than human driver.

2. Better fuel consumption - again, more optimised driving patterns.

3. Better convenience - you can drink and drive(well, car will drive), you can sleep and drive, you can work and drive.

4. Better transport - self-driving trucks don't need breaks, are not likely to be overworked and will likely not block motoways taking over another truck going 2 km/h slower (yey European highways!).

5. Better vehicle utilization - we will be able to reduce total number of vehicles and all the congestion, environment pollution and money that comes with it.

While solution you mentioned really improve quality of life in cities, they do not ultimately solve issues inherent to car transport, where car transport is actually preferred/best solution.

Good public transport systems provide even better benefits for points 1 (accidents involving subways are probably still going to be rarer than for self-driving cars), 2 (you don't even need fossil fuels for subways / trams), and 5. Arguably you don't lose much on point 3 either, except that you may have to walk a few minutes to/from the nearest stop (which is probably good from a public health standpoint anyway).

I agree that this only applies in cities though, and public transport may not help much in rural areas.

> ...still the whole thing works like a charm.

I find dense cities don't work for everyone. Car towns are still nicer for the disabled, the elderly, and families with young children. Especially if stairs or snow enter the picture.

There is probably a way to build dense cities that work for people that need wheels and canes to function, but rebuilding cities to provide ramps, elevators, wide doors, and secure motor scooter parking is also expensive. There is also a last-mile problem for public trolleys and buses to be both available to the disabled and efficient for general use. That's why I think self-driving vehicles have a lot of promise. It probably makes more sense to have a larger fleet of smaller busses and then another set of small busses or vans for people who need a ramp up into the bus, locking mechanisms so their chair doesn't roll around, etc.

I'm going to have to firmly disagree on all points.

Car towns are the worst for the elderly. Most elderly don't feel comfortable driving but in the US, they're all but forced to do it to escape the house. Anywhere else, they can get out and socialize. Getting physical and mental exercise. You're also likely to find nicer single-floor homes (good condos with elevators) rather than "old people warehouses". As for snow, do you really want drivers with failing eyesight and reaction time to be driving on snow and ice? Or even shoveling out their car, for that matter?

I have young children. How many articles have we seen here about kids walking to school or childhood independence? Being car-bound makes that extraordinarily difficult or impossible. At this point, I struggle to think of an advantage of a "car town" for kids. Less time spent getting dressed in the winter, maybe? I suppose when they're older and even more "free range", I'll have to worry about how far they've gotten on the bus or train without me. I'll consider that being a victim of success though.

> Most elderly don't feel comfortable driving but in the US...

I'm talking about people who can't do stairs anymore. They're not walking a few blocks or driving cars. They're taking shuttles places or taking trips with the family. Shuttling the elderly around is cheaper and simpler in car towns. Scooting around NYC or downtown SF with Grampa in his wheelchair is rough. Rougher still when it's cold, raining, or hot. It's often bad for their health, even.

Car towns, maybe just because they are newer on average, are easier to deal with.

* basically never having to illegally park just to function

* car towns have plenty of bathrooms and indoor places to stop and relax

* elevators and handicap accessibility is much more common; dense cities are lousy staircases, steps, and narrow aisles

* keep gear in the trunk (changes of clothes, extra diapers, umbrellas, ponchos, picnic lunches, etc.)

* buckling a tired kid in a car is safer and more convenient than dragging them across intersections

* a screaming kid in a car is nicer all around than a screaming kid on a train

* getting around a dense town with luggage is rough: cabs don't stop, big/extra suitcases on buses is rough, and there are stairs everywhere

The math changes, of course, if everyone in your party is old enough, healthy enough, and responsible enough to carry their own gear. Some families with kids are in this category. But most families aren't that way, at least through different seasons.

Tens of thousands of people die every year in vehicular incidents in the US. Who knows how many more are significantly affected health-wise and otherwise in their life from such incidents. With self-driving cars, those numbers would plummet. It has nothing to do with your other claims, however I would say that it fits the Americanized "because we can!" fantasy.

Denser cities don't help that much, because you'd still have the massive scope of inter-city travel. The whole of Great Britain is only a couple hundred miles long in its longest axis, so it's not an applicable model.

As an American one of the things I'm optimistic for is that cheaper, more available per trip transportation makes more housing viable for people that don't want a car.

If that happens, it's a way out of the car trap that we are currently in, directly, but also because more people living without cars will lead to more infrastructure for people living without cars.

> better public transport

The Seattle Metro area is working on that. Building out just a few mass transit lines here, not nearly enough, is going to take until 2050.

By which time the built lines will already be far over capacity.

> more cycling

I love cycling, but Seattle is built on hills. A lot of them. Very steep ones. Also it rains, rather frequently, up to 150 days of the year.

Electric bikes help solve the hill problem, harder to solve the issue of everyone going around smelling like a wet dog. Tech companies may offer on site showers, but a lot of other companies don't care enough about their employees to bother.

But even with all that aside (I am a huge fan of mass transit), let's take a look at driving between metros. Right now mass transit between metros typically involves going from one city center to another. So in a world of amazing quality mass transit, it is a trip to DT of your home city, a ride to another city, then a trip to your final destination.

Replace that with on call self driving cars. Picked up at home, and get dropped off close to your destination in another city. Within that city, and within your own city, use mass transit.

With the light rail Seattle is building out, my commute from one of the surrounding cities consists of driving to a park and ride transfer station (because even though I live on a main thoroughfare, mass transit isn't coming anywhere near me in my entire lifetime!), then taking mass transit in to the city.

If I avoid rush hour, my travel time to city is 15 to 20 minutes.

An autonomous car picking me up and taking me into the city makes a lot of sense. Indeed, I'd prefer if all the money (and decades of work!) being spent connecting the cities around Seattle was instead spent building incredible high quality mass transit within Seattle, and the surrounding cities.

At that point, serious change to the layout of cities could happen. Transit could get ran everywhere, huge parking lots removed, lots of housing built close to work, and people who want to would be able to live close to work, helping ease congestion.

Self driving cars make that possible.

The technology that is being built in America needs to be built up around the, unfortunate, fact that we have suburbs. You cannot relocate the majority of the country's populous. There are 700k people in Seattle, and 3 million in the surrounding urban area!

We can have a hundred millions of self-driving cars dumping the sub-centimeter LIDAR data of surrounding to OpenStreetMap at any given time.

Come to think of it, I see zero reasons why you won't own measurements of your self-driving car, and choose to upload them. Yes corporations may try to pass it as their intellectual property, we should fend them so hard they roll back into XIX century.

Cars are going to lock down their LIDAR systems, so the best one can hope for is that the companies that control that data are also friends of OpenStreetMap and will pass the data on. I’m not optimistic that the individual owners of cars will be able to read from the LIDAR bus without – at best – voiding their car’s warranty.

> I’m not optimistic that the individual owners of cars will be able to read from the LIDAR bus without – at best – voiding their car’s warranty

That's one consumer protection law away.

And no, I don't believe in "being friends with OSM". I believe in owning your data.

> That's one consumer protection law away

“If you buy the car you own the data stream, and must be able to keep it private; read, modify and re-distribute it; and sell it” is a good idea.

> I believe in owning your data

Gentle political note. This sounds more radical than it is or than your argument needs it to be. The first suggestion stands well enough on its own.

Not sure open street map really wants to be in that business...

The thought provoking contrast presented in the article is camera derived sparse maps with lane lines, road edges, signs etc. vs lidar derived dense detailed models of the whole environment. Any thoughts around why the former is not enough? If it is enough it would certainly be preferable, since smaller in size and likely easier to maintain.

My guess is that the more detailed, dense map allows the car to precisely localize itself by using its view of the world to match against the map, yielding a location more precise than GPS (and for when gps doesn’t work); eg ICP point cloud matching/alignment/registration. More sparse maps would give you a less confident match, and you’d have to convert what the car is seeing to the sparse form.

The former is enough, but having the latter makes it much easier.

Bumped into this comparison of Google Maps and Apple Maps and it is excellent.


Motivated by a road that was on one map but not the other, I built this: http://comparemaps.drona.ro/

I've been wondering when Google (or someone else) is going to do the same thing for pedestrian routes and locations. There are plenty of highly-trafficked places where GPS is unavailable or not precise enough, and mapping WiFi stations evidently isn't enough to fill the gap. Of course the problem could, and should, be attacked by ground-based radio systems * , but 100% coverage by such systems certainly won't arrive anytime soon, and it wouldn't be like the tech companies to wait for it. You would have to take your phone out to get a location fix ... but that likely won't be the case whenever AR glasses really start to appear in public, and the techcos will also be happy to stick cameras on your clothes and luggage if they and you can get away with it.

* Come on Transport for London, it's high time to show some leadership here.

Niantic's Ingress kind of demo'd an early approach to this idea of pedestrians doing map data collection -- http://simplify360.com/blog/ingress-how-google-is-gamifying-...

Quote: > The game is always prompting the player about his location. In this regard, Michael Carney from PandoDaily said, “To capture a Portal, and harvest the “energy” contained therein for his respective team, a user must physically go to a location and check in. additional energy is available by travelling specific walking paths, bike paths, and inner-city routes dictated by the company, all while the user’s Android device is transmitting GPS and accelerometer data. In some cases, the user will be required to photograph locations or objects along those routes.”

Quite literally trading magic beans to punters, noice. For high-value pedestrian-only areas it would also be quite viable to send out Google employees or gig-workers. Likely with dedicated hardware, which wouldn't have to be very big or alarming: a consumer WinMR headset is already doing a first approximation of the kind of scanning you'd want.

They announced something for this at IO last year: https://techcrunch.com/2017/05/17/google-has-an-indoor-posit...

Aha, thanks! Time to start the countdown to self-driving suitcases so.

Learning how to look at a map, look at your surroundings, and then look back at the map to figure out where you are is apparently a lost art.

Do you primarily mean high density walking areas, or essentially everywhere that you can freely walk as a pedestrian?

Naturally you'd prioritise things like all the exits from major metro stations in tier-1 cities. Further down the curve, if it's viable to maintain a map of every public vehicular road in the USA then it would surely be similarly viable to map all the major public streets and squares in the centres of the world's tier-1 cities; likewise the major public interior spaces in their public-transport buildings. So that's your MVP; then obviously you'd go as far further down the long tail as technology and business make viable (and regulation permits).

Basically it's just a combination of the SLAM tracking being successfully used already in VR and AR headsets like Windows "MR" and Oculus' Santa Cruz with the usual data-consolidating ways of Larry's Basilisk. I have a suspicion this may be the intended final purpose of all the nice Project Tango research.

Obviously you could also use the database for things like photo location-tagging. Similarly, smartphone car navigation apps would want to be able to be able to use the data being collected for self-driving cars, too. You could also imagine self-driving street-sweepers taking advantage of the pedestrian-area data (or indeed helping to harvest it).

In Europe in particular (and I'm sure elsewhere) a lot of pedestrian areas are very narrow. The lack of precision is a big problem. I have not had same number of issues in the US presumably because so many pedestrian areas were built with cars in mind (at least at first).

Or just avoiding roads with no sidewalk. This has improved by sometimes Google still wants to send me via roads that are not suitable for pedestrians.

Interestingly it's the opposite for cyclists. There, Google chooses "quiet" routes, sometimes sending me via paths where cycling isn't allowed. Could be a London mapping issue though.

High density, high intersection pedestrian areas. I’ve gotten so lost in Japan metros before, for example.

No pois, no maps.

Google is unfortunately still the winner here.

well all my POIs are in gmaps and I have the app installed everywhere. No idea why I wouldn't want them to stay at the top. Push them please, so they improve their software, but winning they should.

Nobody wants to let Google win again.

I do. Google greatly enhanced my life, esp. Driving and directions.

I was an adult BI and speak from experience

Just last week Google maps directed we take a turn going down the wrong way of a one way street that was already clearly marked as one-way on the map- yeah, I'm not too worried about even Google's dominance in this area.

Also, Apple maps in my area has been quietly, but steadily improving. Still not as many POIs as Google, but nicer interface and turn-by-turn (IMO).

Where did the one way street thing happen exactly? I find it very interesting that it can do that, it implies a disconnect between the rendered data and the data used for routing.

It could also be a simple routing bug. You cannot know.

It's also very frequently people being in 'walking' mode without realizing it

I'm surprised the accelerometer doesn't question its input when I appear to be doing 75mph on foot.


Waze: https://arturrr.com/2018/02/19/navigation-apps/

I've experienced something interesting a few times in high-traffic situations (closed highways/etc) where byways produce time estimates which are instantly implausible as many people attempt to take them. It's interesting to think of methods Waze could use to predict optimal paths while accounting for its own effect on traffic.

Get me a waze _lite_ and I’m sold. But not current waze.

Wade takes me from a to b in any country without effort and faster than any other.

Wade it is. Yes the chat function is largely useless the report incident function is cool!

> faster than any other

Are you sure? Someone just did an interesting analysis (admittedly, very small data space but a long-term test), found that Waze claims faster routes than they typically achieve.

Update: Ah, @dogecoinbase linked to it: https://arturrr.com/2018/02/19/navigation-apps/

I am here providing an anecdotal story.

Zurich Switzerland -> Austrian Alps

Heavy traffic, incidents along the road, gridlock ...> Waze got us there FASTEST.


Competed against iPhone & Google Maps.

Same way return drive. ... So two data points I guess.

Austrian Alps -> Zurich

Plus ... I use it daily... So three data points then.

It's entirely possible you're right, but I don't see how on any individual trip you can assert any solution achieved the best time, unless you have 3 different drivers to take each proposed route (and even then, there's enough variability to make it suspect unless you do it over a long enough period of time).

Well, when this happened than really everyone no need to worried about to lose their home in the world.

It's not about mapping. It's about collecting data and training their machine learning algorithms.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact