Daily Placebo

  • Wednesday, November 19, 2008
  • The Ants Screw Me Again.

    Okay, so I've got a bit of history with ants. Honestly, these dudes are annoying, right? Well some kinda researchers think that ants may be the solution to some of our traffic congestion woes.
    When the narrow route, as expected, became congested, ants know to tell each other to redirect their travels. An ant returning from the congested narrow route encounters another ant heading up that route and pushes that ant towards the wider lane.

    Seems like it'd make sense, right? Individuals who've just traveled on a path can communicate with those headed the opposite direction into the observed conditions. Oddly enough this eerily reminds me of an idea I postmarked on 2/21/07. Take a look:

    Wireless Real-time Itinerant Social Traffic System

    A mobile system that uses social information, made publicly available, to
    compile useful individual results for a single entity of the entire
    social group.

    The system would harvest speed, congestion and general flow
    information from a vehicle's on-board sensors and create a recent
    database track with geographic locations, times and infrastructure
    conditions. No optical or radio traffic monitoring equipment (to gather data on opposing cars
    without WRISTS installed) is required, but may enhance the overall
    community performance if present. As the vehicle proceeds on its route
    it actively searches with a wireless connection for oncoming vehicles
    also equipped with a compatible system.

    When a peer is found the systems trade recent databases,
    merging current road condition information. Each vehicle now possesses
    a complete local traffic map of the area with reliability ratings pertaining to data recency and surety of diagnosis. (some traffic
    patterns are easier to diagnose) Vehicles may also choose to rate
    personally collected data higher, or flag data, so that inconsistencies
    can be leveled out of the network should errors arise.

    The in-vehicle system will then use this information in its on
    screen navigation, creating alternate routes to minimize travel time
    (or another variable at the users discretion) to way-points in the
    infrastructure. The information may be displayed on the navigation
    screen by classifying roads into color groups, or may be used silently
    by the system to reroute traffic.

    Systems will need to be self tuning, calibrating for how long
    data may be used before it is no longer representative and must be
    discarded. In some areas 2 hours and 100 miles of data may be useful,
    while in other areas 15 minutes and 4 miles is the maximum lifespan.

    The system will also learn tendencies of specific roads over
    time for personal use, should a user frequent a particular area. This
    will allow for general prediction of traffic
    flows in routing. So if a thoroughfare is clear at 7:30, but is
    usually full at 8:10, the system can compute ETA to the hot spot and
    smartly disregard "road clear" information that it calculates likely to
    be out of date on arrival. This average road condition prediction is
    not for social use, only recently harvested information is transmitted
    throughout the network. In addition existing traffic broadcast systems or fixed traffic
    counters may be integrated into routing because they mark known
    accidents and closures, but this information is for personal use, not
    socially distributed. This is intended to maintain the purity of the
    network so that only fresh and near-firsthand information is used which
    will eventually result in better local response times than widely
    aggregated reports of a metropolitan area.

    With multiple users an entire area of main thoroughfares can be mapped fairly quickly, depending on traffic speed and transmission range. The more users there are, the more current and ubiquitous the data will be.

    When enough users are enabled, oncoming traffic will act as a conduit for road information to stopped traffic. For example(figure 2): A car (A) in a traffic
    jam can pass information to car (B) as it proceeds in the opposite
    direction. Car (B) will in turn pass information to car (C) which is
    pointed the same direction as car (A) and in the same traffic jam. Car (C) can then use information passed backwards down the traffic jam to decide whether to reroute or to wait for traffic to clear.

    In more advanced situations cars with similar repeated tracks may suggest "bond" may also communicate destination way-points.

    study: minimum number of devices for system to function. Effect of
    congestion avoidance on quality of data, multiple concurrent reroutes.
    Make extrapolated congestion based on routes and reported trends
    public? (figure1: assume traffic is red between 7:12 and 7:17 links)

    Freaking ants stole my idea and sold it to some German Scientists! And you call me crazy when I rant about ladybugs trying to kill me and not being able to trust insects.

    Labels: , , ,

    digg it


    At 6:48 AM, Blogger theKirkness said...

    W.R.I.S.T.S for short!


    Post a Comment

    Links to this Post:

    Create a Link

    « Daily Placebo