Government control and road pricing, and GPS technology

It was suggested in an earlier post that a GPS system with appropriate interface with a car's information system could provide a platform for an intelligent cruise control system, which effectively removes the possibility of the driver speeding. It is quite possible that governments will eventually require drivers to install such systems. (Of course, much of the law enforcement effort will then shift to identifying those cars whose drivers have disabled the system.) However, it's far more likely that drivers will install such systems voluntarily, in order to avoid the very substantial fines that can be racked up very quickly in highly invigilated societies such as Australia and the U.K. (and, to a lesser extent, the U.S.). Alternatively, insurance companies might offer very substantial discounts for drivers who agree to install such systems, as a means of reducing the frequency and severity of accidents.

Once installed in a car, of course, the GPS can also be used as a record-keeping device. Potentially, the GPS unit might communicate with government registration or road pricing databases. Such a system would avoid the need to install physical tolling infrastructures such as overhead RFID gantries, or indeed human operated toll booths such as are still found in parts of Europe and North America. Clearly, citizens would be tempted to hack into such units in order to minimise road charges accrued. Data integrity and security therefore promised to be a big part of the future of GPS technology research.

Such ideas also lead to considering the possibility of GPS be used context of a dynamic road pricing pricing system. Citizens are sure to find road prices which change daily according to conditions highly annoying, unless they can delegate the task of calculating the cheapest possible route on any given day to a GPS system. 

GPS communication, learning and interface

Two-way communication and learning

In my view, the biggest improvements to GPS systems in the next few years are likely to be in the areas of learning, AI and communications. I am continually frustrated when my GPS system (a Garmin Nuvi 255W, since you ask) wants to send me along congested city streets, instead of taking the longer but quicker way on the freeway. What frustrates me is that the GPS system doesn't learn how quick or slow those streets tend to be at particular times a day. A system that learned and optimised as it went along what streets, or even what kinds of traffic features, tend to speed up or slow down traffic would be immensely useful. For example, in Melbourne, tram routes tend to be hopelessly slow, a fact that any human driver knows, but one that seems to elude my GPS unit. 

We all know that GPS units are able to receive information from dedicated traffic channels, indicating current road conditions. It should be possible for GPS units to communicate traffic information "back to base", so that the traffic channel effectively receives aggregated real-time traffic information from GPS users themselves. Over time, the sophisticated model of traffic flows over the entire road system could be built up quite quickly. Such data would be extremely useful to road, public transport And housing planners, as well as to the oil industry and to statisticians. 

Such information, of course, could also be used by map providers to verify their own map information and to accommodate short-term phenomena such as roadworks detours, and also to identify new roads in real time. Care would of course need to be taken to avoid adding private driveways to the model public highway system.

GPS unit as a source of in-car regulatory information

Gradually, it seems inevitable that more and more public datasets will be integrated into the map providers' POI (point of interest) databases. One dataset that would be particularly useful would be public on-street parking information, in other words, for each street, and for a given time, what the maximum parking period is and what are the conditions attached. Armed with such information, the GPS unit could navigate the driver to the nearest on-street public parking which is available at the applicable time and for the appropriate duration. 

Similarly, the GPS system could display the location of all known traffic lights and, indeed, assuming communication between the GPS system operator, and the public traffic light control system, it would eventually be possible to display in real time the state of traffic lights even before they become visible to the driver. The GPS system could also display a recommended speed in order to arrive at the lights in a green state. Clearly, the GPS system is not the only possible means of displaying this information, but it is the most likely candidate for this job.

Interface issues

Voice recognition

Operation of a GPS system can be a dangerous distraction, if the results of a British survey are to be believed. I can in fact believe it, and have experienced first-hand potential distraction encountered while attempting to enter a route and drive at the same time. I dictate this blog using Dragon NaturallySpeaking dictation software. Version 10 is not perfect but it's a lot better than the layperson might expect. My guess is that GPS systems will feature voice recognition in the very near future. the main challenge will of course be accurate recognition of street names and other points of interest across a range of drivers who might wish to use any one system. Assuming these challenges can be overcome, interacting with your GPS via voice will radically improve your safety.

Head-up display

One problem that most GPS users encounter is the difficulty of finding somewhere to place the unit, which does not block the driver's view of oncoming traffic but which remains within reach. Assuming that voice recognition can remove the need to have a touchscreen within reach, the prospect of integrating GPS information with a heads up display (HUD) becomes more real. Already, some cars are available which project basic GPS information (such as next turn) on to the windshield. In future, the windshield will resemble a computer screen, on various parts of which is projected information such as speed limit, next turn, and traffic light and parking information as discussed above. At night, it might even be useful to drivers on dark highways to have projected an image of the road ahead in correct perspective.

In my next blog, I'll tackle GPS's potential use in dynamic road pricing.

Automotive GPS integration

We've most of us seen those high-end cars with integrated GPS systems, centrally mounted on the dashboard. Integrated satellite navigation, as it is often called, remains a very expensive option available to the automotive purchaser, often adding thousands of dollars to the list price. By contrast, third-party GPS systems can start at some 200 Australian dollars. 

However, right now there is no way to get third-party GPS to talk to the average automobile, rather like the situation until very recently where car stereo systems would only accept input from MP3 players such as iPods, by way of radio adapter gadgets. 

Think of the technical possibilities if your car and your third-party GPS system could talk to each other easily:

• Calibration: did you know that your car's speedometer is likely to exaggerate your speed by roughly 4 km/h at 100 km/h? Presumably, motor manufacturers do this, in order to save you from your own propensity to break the speed limit. Whatever the reason, it's annoying for those like me who wish to drive at the precise limit, whenever it is safe to do so. For this reason, I normally drive with my GPS unit on, even when I know where I'm heading, as the GPS unit has an extremely accurate speedometer. Similarly, those with cruise control option in their car would obviously like to set the precise speed to avoid speeding fines while maximising speed. given the superb accuracy of GPS unit, it should be possible to calibrate your car's speedometer at will, and even to run a digital speedometer, directly off the cheapest unit. 
• Dynamic cruise control: modern GPS units model that currently applicable speed limits for every location. Using such information, as well as the GPS's speedometer and altimeter information, it should be possible to design a cruise control system that is far more sophisticated than the existing offerings. Such a cruise control system would dynamically adapt to the changing terrain, preventing over-and under-shooting of the correct desired speed. More excitingly for the law-abiding driver wishing to avoid speeding fines, such a system could readily be configured so as never to break the applicable speed limit.
• Trip computer: given sufficient two-way communication between your car and your GPS unit, either one should be able to perform all the functions of the most sophisticated trip computer, including current and historical fuel efficiency data as well as distance to empty, distance travelled, etc.
• Automated indicating: your GPS gives you turn by turn instructions, no reference to your selected destinations so why shouldn't it also control indicator of your car obviously, it should be possible to override the automatic indication whenever you choose to do so.
• Automated steering? This is a no-no right now and for the foreseeable future. Given that GPS is currently accurate only to 10 m rather than to the 1 m theoretically possible (due to encryption of the signal by the US military), automated steering controlled solely by the turn-by-turn and lane navigation systems of your GPS unit would be hopelessly unsafe. Nevertheless, GPS systems may one day, provide at least some of the input into automated steering systems for the highway, just as they have done in the DARPA car challenge. it should not be forgotten that the European Union is planning to launch the Galileo system of navigation satellites, which promise greater accuracy than the existing GPS system. By then, some of the more daunting problems such as automating obstacle avoidance, keeping maps sufficiently up-to-date in view of roadworks, etc, may have been solved. 
  

What seems to be lacking is agreement on a communications protocol which would connect third-party GPS systems with a range of automobile marques. What could be holding this work up? Is it a case of a battle of incompatible standards, similar to what we witnessed over the last few decades in fields such as video cassette recorders, DVD burners and high-definition DVD players? Perhaps, but my guess is also that car manufacturers are not playing ball for their own reasons, either simply through resistance to really getting the best out of this new technology or, perhaps more likely, because close integration with third-party GPS systems would jeopardise the very fat margins they are currently able to charge for in-car satellite navigation options.

What is needed, probably, is for one or more innovative car manufacturers to team up with one or more GPS manufacturers to come up with a communication protocol. GPS conductivity, could be a real selling point for car factories, just as iPod connectivity has become more innovative manufacturers such as Hyundai. Ideally, the protocol settled on would be an open standard rather than a proprietary standard, but it probably doesn't matter too much, as long as some protocol is. 

In my next post, I will continue on the theme of possible improvements of GPS technology more broadly, including using GPS technology as a platform for in-car display of regulatory and other information.

An Integrated Bushfire Information System (IBIS)

Many readers will be familiar with recent wildfires in California, in Greece, and most recently the massively destructive bushfires in February 2009 in Victoria, Australia, where my family lives. On so-called 'Black Saturday', 7 February 2009, 46C temperatures and 100 kph-plus winds fanned wildfires that, according to evidence at the subsequent Royal Commission of inquiry, released energy equivalent to many thousands of atomic bombs, killing 173 people and destroying over 2,200 homes. 

The recommendations below are drawn from my submission to the Royal Commission.  The submission gives the background and a critique of the existing internet publication channels for bushfire information, which are very inadequate. 

The submission argues that the cost of extreme bushfire events, and the increasing probability of such events given the reality of climate change, necessitates a step change in the level of investment in information technology and specifically geo-Informatics to help manage the risk. Specifically:

1. The aim should be the construction of an integrated bushfires information system (IBIS) that would be accessible and used by the public and authorities in managing their bushfire risk. To the fullest extent possible, members of the public should be able to access the same information as is made available to the authorities.
   
2. The aim should be for a portal or Internet page which is as timely, comprehensive and granular as humanly possible, and certainly far more timely, comprehensive and granular than the information available on Black Saturday.
   
3. The information displayed on the portal should be as close to real time as possible. It should be clear which agency is responsible for each item of information. To the extent possible, in order to maximise the freshness of the information provided, automated sources of information should be relied on, including the bureau of meteorology's databases, the outputs of satellite systems such as GeoScience Australia's Sentinel system for displaying satellite information, etc.
   
4.  The basic medium or matrix for the presentation of information should be the interactive map. The technology could be as simple as that used by Google maps, but preferably would be purpose-built. Wherever possible, information should be as accurate as possible given modern GPS units’ capabilities. To this end, all firefighting units should be equipped with GPS technology and all reports should be automatically geo-tagged wherever possible.
   
5. The portal should be capable of displaying integrated information from all relevant agencies (State and federal), but particularly from firefighters, roads authorities, police, the meteorology bureau, and sources of satellite and aerial surveillance imagery.
   
6. The meteorology bureau or other appropriate expert organisations should be tasked with the automated production of real-time critical fire path projection mapping. This would show the projected path of the fire front of critical fires given known variables such as fire location, windspeed, temperature, precipitation and the like.
   
7. Residents and possibly other members of the public should be able to post sightings or reports of fire in an unverified layer of the map. In order to minimise the risk of vandalism of the site by scurrilous reports, suitable safeguards such as logging IP addresses or perhaps even establishing a system of pre-screened volunteer fire reporters should be established.
   
8. The investment must be both in terms of equipment and expertise. The authorities must commit to sourcing highly expert specialists in what is a cutting edge field, recognising there may be a need to pay over the odds for such expertise. All relevant staff should receive training in the use of information technology and geo-Informatics in the reporting of fire and in the communication of relevant details of the public. 
   

First Post - What's it About?

The rationale for this blog is very simple.  I'm often coming up with ideas for new products or refinements of old ones, but I'm not an industrial designer and I don't have a great deal of time, patience, or money to develop, refine and commercialise these ideas. But I would still like these ideas to be 'out there' and helping to contribute to the creation of better products in the real world.

 

When I was a boy, I would love to invent new machines and devices, and I spent seemingly hundreds of hours drawing plans for new inventions. Often these were fairly crude doodles, but sometimes these were more elaborate drawings and plans, and occasionally these "inventions" were really pretty advanced for a young boy. For example, when I was about 11 or so, around 1977, I drew up designs for the carriages, tracks and points for a maglev train system, at a time before I had ever read about magnetic levitation train technology. Again, around this time, I "invented" a means of aerial navigation, relying on radar scanning of the underlying topography and matching those scans with computerised countour maps. This technology, I later discovered, was deployed by the USAF in its nuclear cruise missiles. (Incidentally, one drawback of the technology, which has been superseded by GPS, is its unreliability in deserts and snowfields conditions. )

There will basically be two categories of idea forward on this blog: 

• Firstly, "half baked" ideas for new products or new software. As almost everyone knows, there is no copyright in an idea, so any reader is free to take this idea and try to refine, operationalise and commercialise it. It would be great if those readers would, via the comments section or by becoming contributors to this blog, keep their refinements to the basic idea open source and in the public domain. However, at some point the investment made by a reader in refining an idea into a product may well require secrecy in order to maintain enough novelty to ground a patent application. To that extent, there is no intention to try to enforce openness on readers.  

• Secondly, actual designs for inventions, possibly to seek comments and refinement from readers. If any readers wish to contribute to or commercialise these drawings, I would be very interested in entering into discussions with them, via the comments section or directly. Once again, the idea is to have my ideas put into practice "out there" in the real world. 

Well, enough preliminaries. What ideas do I have have up my sleeve? Read my second post to find out.