Gate to Gate

Over the last couple years, I have written about many bits of technology that aircraft use. I really haven't discussed too much what bits are used when, and for what. This post, I will try and tie all the items and their use into a comprehensive post. I'll use a commercial airliner (Part 121) for the discussion, both because they typically use more technology, and because that is what my background is. I will also base most of this on flights in the US, to keep it simpler.

We can start a few hours before the flight actually leaves. As the flight approaches it's scheduled time to leave, a group of folks have started planning for the flight. Sure, there is network planning, they set up the schedules and try to be sure the flight will make money, and provide continuity, and such, but that is usually done months in advance. There are also the simulators that the pilots use for proficiency checks, and training, but that is on going and not related to a particular flight, but important none the less.

Dispatchers and meteorologists are considering the situation between the origin and the destination of the flight. The meteorologists are generalists, looking at the weather over the country, where dispatchers are more concerned with the weather along the route between the cities. The dispatcher needs to consider the situation at the specific airports, for runway closures, and other challenges unrelated to weather. There is a tool called Collaborative Decision Making (CDM), where the dispatchers work with the other airlines so everyone can utilize the airports and airspace as optimally as possible.

The dispatchers can use CDM to look for areas to avoid when selecting a route between the two cities. If there is a bad thunderstorm along the optimum route and all the other airlines are avoiding it to the south, the dispatcher may pick a northern route to stay out of everyone elses way. Once the dispatcher selects a route, they need to build the rest of the flight plan. The dispatcher will build a flight plan using many tools. The dispatcher may allow the flight planning engine to select routes, or the amount of fuel. Depending on aircraft maintenance situation, and MEL deferrals and such the flight planning engine can accurately predict the fuel burn based on weather and route.

Once the dispatchers are happy with route and fuel selected, they will file a flight plan. The flight plan will be filed with the Air Navigation Service Provider (ANSP) for both the origin and the destination. For the US the ANSP is the FAA, in Canada is is NavCanada, and in the UK it is NATS. The ANSP handles all the RADAR and air traffic control (ATC) functions. The flight plan will give the ATC controllers a heads up on what the aircraft was planning on doing once in the air.

As the pilots get to the aircraft, one will typically do a walk around of the outside of the aircraft making sure the aircraft looks safe and no damage is visible. The other pilot will typically go to the cockpit and begin setting things up. There may be a the initialization of the FMS, maybe a RAIM check of the GPS, and entering the flight plan prepared by the dispatcher into the FMS. The dispatcher provided flight plan will usually include weather information for the route, and any other non weather realted information for the route (IE ATC changes, etc), The flight plan will also contain fuel and time information that the pilot can double check, insuring the dispatcher hasn't made any mistakes.

When the pilot know the fuel situation, they may confer with the fueler to adjust any fuel amounts to be put on board. The pilot will also need to know how many bags and passengers are on the flight, so they may make proper weight and balance calculations. Some airlines have a load planner who takes care of the weight and balance, others still let the pilot take care of this. Depending on the aircraft, it may be necessary to have a person dedicated to making the load calculations.

As all the passengers are seated, and the pilot is about to move the aircraft, they will ask for permission to move. There may be a ground controller dedicated to the gate area, and there will need to be taxi clearances and such needed from them. Other airports everything is controlled from the tower, and any movement must be cleared through the tower controllers. An ACARs message may be sent requesting the Pre Departure Clearance (PDC), that will be a version of the flight plan sent to the ATC with any ATCneeded changes to the plan. The PDC will also contain the code the pilot needs to enter into the transponder. The pilot must acknowledge receipt of the message.

After the aircraft is taxied to the runway, the pilot will ask the tower for the final airport clearance, by announcing "ready for takeoff". Once the ATC controller gives the pilot final instructions the pilot can access the runway and start the takeoff. The ATC instructions will be the route the pilot should take to get from the runway to the beginning of the flight plan route. Every bit of the instructions and plans for the takeoff are there in case there is a failure. If there is a radio failure either from ATC or the Aircraft, the instructions given are good enough for the pilot to take off, fly the planned route, and approach the destination. It is a safety situation, should the plan be the safest and most expeditious way to fly the route.

Once the aircraft is above about 300 ft, depending on the  airport, the aircraft will appear on RADAR. The first RADAR that will show the aircraft is the TRACON, who will control the aircraft after tower hands off the aircraft. The TRACON controller will control the aircraft until it is more than 30-50 miles from the airport. The TRACON will hand the aircraft off the enroute controllers who will control the flight until it is 30-60 miles from the destination airport. The RADAR data will be collected and sent to the FAA command center for others to view, and use the ADSI information. As we move into NextGen, there may be more ADS/B position reporting, instead of RADAR.

Once the aircraft is on the route, the pilots will typically engage the autopilot. The autopilot will help maintain the route of flight, altitude and throttle settings to insure the aircraft flies the route planned, and uses the fuel planned. The pilot must monitor the autopilot to be sure it is engaged, and doing the right thing the whole flight. Occasionally pilots will hand fly the aircraft, for practice. Once in a while the autopilot will fail, and the pilots must had fly the aircraft. The systems in the aircraft are designed for certain reliability levels.

During the enroute portion of the flight, there may be messages the pilots need to send to the company operations center. The pilots will usually send a text message over ACARS if they don't have a lot of urgency to the message. The pilots also have an option for voice communication using company assigned frequencies. If there were to be a medical emergency, the voice communications will be used, if a pilot is looking for a weather report for 400 miles ahead, they will use ACARS.

As the aircraft gets closer to the destination, ATC will typically begin having the aircraft start to descend. Newer approaches follow a continuous descent profile, where the pilots set the throttles to idle at altitude, and basically use the potential energy to glide the aircraft to the runway, reducing noise, fuel burn and pollution.

Current approaches typically are designed for the aircraft to provide it's own guidance. That is there are airs on the ground (or satellite) to provide the aircraft the information it needs to know where it is, and fly to the runway. Features like DME and ILS radions are on the ground, and GPS satellites are in the air.

The enroute controller will hand the flight off to the TRACON controller about 50 miles from the airport, where the aircraft will be below about 10,000ft. The TRACON controller will clear the flight for the approach that it will use to get to the airport. At about 5 miles out, the pilot will be told to contact the tower, and the tower and the pilot will make the final checks and be cleared for the runway to land on. Once on the ground, the pilot will talk to the ground controllers to get to the proper parking area, and maybe a gate controller for certain airports. Once the wheels are chocked, and the engines shutdown, the pilots are mostly done with the flight.

Yes, there is a bit of technology going on between each gate, and a little before. Ever think about that before.

UAT or 1090ES?

If you are considering ADS/B, there is a choice to make. Do you install a Universal Access Transceiver (UAT) or the Mode S transponder that has an extended squitter (1090-ES)? It all depends...

What country are you in? If you aren't in the USA, then the choice is pretty much made. The USA offers the option of a UAT. The rest of the world needs Mode S transponders for ADS/B installations.

If you are in the USA, and you mostly fly above FL180, then the choice is pretty much made again. The FAA doesn't allow aircraft flying above 18,000ft to use the UAT. It just makes sense to get the 1090-ES transponder that will do Mode S if you want take advantage of ADS/B and fly about FL180.

The UAT transmits and receives on 978MHz, the 1090-ES transmits and receives on 1090MHz. The ADS/B system will allow all participating aircraft to see each other. If the two devices work on different frequencies, how does a 1090MHz transceiver see a 978MHz transceiver? The ground stations will repeat the 978MHz messages on 1090MHz, as well as repeat the 1090MHz message on 978MHz. The ground station will also show both messages on the "RADAR" scope, so the air traffic controller knows where everyone is.

The FAA separated the two systems for a couple reasons. The 978MHz devices can handle more data (has more bandwidth), so more aircraft in a concentrated area will work without overloading ground stations or other aircraft. The 1090 Mode S transponders are already on the larger faster aircraft that are flying higher, so the expense should be minimized (I am repeating the FAA here, in reality, most operators will need to replace the transponders they have to get the extended squitter feature).

The UAT's are even more useful, since the FAA will broadcast extra information. The two extra messages that the FAA is broadcasting are the TIS/B and FIS/B. The 1090-ES system will get TIS/B, but not FIS/B.

TIS/B is Traffic Information Service-Broadcast, where non-ADS/B equipped aircraft will show up on the aircraft display, similar to ADS/B equipped aircraft. The ground station will broadcast the position of aircraft that are only visible on RADAR. As a pilot, you will be able to see more of what the controller sees.

FIS/B is Flight Information Service-Broadcast. Flight information includes weather, and aeronautical products. While XM provides some weather, that you must subscribe to, the FIS/B is free to everyone. The XM product may have additional information, or be more timely. The FIS/B data is what the FAA will be looking at, including potentially air traffic control. The aeronautical products appear to be weather like items, such as NOTAMs and SUA status.

Exactly what device to get will depend on the capability of the chosen display. Many of the MFD manufacturers will take either device for input, the displayed information may help make the choice. Some will show the weather RADAR information in great detail, others will show it blocky or not at all. Over the next couple years, the MFDs are sure to get better.

Should you wait, or should you buy today? Today the ADS/B MFD technology is being developed. Over the next 5 years, the technology will surely mature. Having ADS/B in on a tablet computer will allow a pilot to get their feet wet, sooner. By 2020, most aircraft will be required to have ADS/B out, which probably means, unless someone builds an under $1000 solution to ADS/B out only, most aircraft will be equipped with ADS/B in and out.

Can you get rid of your transponder once you have ADS/B? No, the Mode/C component will still be needed for RADAR service and TCAS for non-ADS/B equipped aircraft.

It'll be an interesting couple years going forward. What do you think?

Here We Go Again

We get to get the threats of the "sequestration" again this fall. Of course we knew about it for most of the last couple years, and who has done anything about it? How about that, the only one who could do something, is congress, the same ones who chose to ignore the trouble.

Congress will play the same game of chicken again this weekend. Nothing new there.  Something will be the pawn, this time it is Obamacare. Fine, whatever, I really don't think we need congress messing with health insurance anyway, I can't see how they can make it better. (this conversation could get recursive about here, so I won't get into congress and health care).

The trouble is, the game of chicken costs everyone more money! Obamacare is in progress, and states have put money into making it work. The money is in the federal budget already, and could be funded. If congress tries to unwind Obamacare, well there will be money wasted by the states, and the fedral government that are working toward the Oct. 1 deadline. Corporations have their HR software ready, the rules were set over a year ago.

So they decide to boot the government out on Oct 1. Well, those government employees just go home, and watch daytime TV and don't get paychecks. Hmmm... they don't buy cars, and limit grocery purchases, and that ripples down to the Best Buy store, and the stock boys at the Albertson's. Yes the whole mess puts a strain on everyone, depending on how long things go on.

Lets see, air traffic control is in play again. Will congress like having their flights delayed? Maybe once, or maybe they have things rigged to not be personally affected this time. If they are delayed, you can bet congress will make a strong effort to adjust budgets to make that part less painful.

How is traffic affected by controllers?

As I wrote in my April Sequestration post, controllers are certified to work a limited section of airspace. If all the people scheduled to work a day can't cover a sector, then planes can't fly in that sector. There are also rules about how long a controller can work a sector without a break (bathroom and lunch breaks are not optional!).

Airspace in IFR conditions is positive control, or only one aircraft can be in a part of airspace until it is known where that aircraft is. I've flown IFR to Livingston Montana in a small airplane before. Livingston is in a valley, and unless you are over 12000ft up, there is no RADAR coverage, and there is no tower at the airport. Once I left Billings MT RADAR area, I was cleared for the approach into Livingston. That meant the controller gave me all that airspace, until I told her that I was on the ground, or back into some other controlled airspace. The controller had to give me exclusive access to that airspace, since she couldn't see me, and she wouldn't know where I would be relative to another aircraft.

If towers are closed the enroute controllers will give the aircraft the airspace around the airport, starting sometimes over 50 miles out. That means only one aircraft can be in that 50 mile area until it is on the ground, or reports that it is going somewhere else. With the tower open, the spacing can be 5 miles between aircraft. With the tower closed it maybe 50 miles of spacing. You can get 10 aircraft on approach with the tower open, and only 1 with the tower closed. Things slow down really fast like that.

Airplanes don't care if the towers are contract or FAA run. If there is proper staffing, then the pilot can fly closer to other aircraft. The aircraft won't have to hold or divert if there are enough people at the tower watching aircraft. There are towers that handle very little traffic, and the staffing could be reduced, and the FAA has tried, but usually some congress person has a special interest in that tower, and it stays open for many hours with few operations.

Again, as flights get late, the second shift or the midnight shift get to work overtime. Congress insures the money saving route actually costs more. If they would just do it right, and leave it alone, they could save real money. You and I have to live within our budget, and changes may cost us money. Congress ignores that whole thought, and let money get spent, knowing that in a year or two they will have to deal with a situation. Waiting until the weekend before will insure things will go bad.

We need a smaller government,  and we could start the cutting with congress.

P.S. Congress will continue to get paid during the shutdown. It is also likely that once the budget is passed, and funding is restored, all the government employees will get retroactive pay. 

Transponders

Surveillance RADAR is very useful, and can be augmented like GPS. RADAR alone can only tell range and azimuth. If the transponder is added, the range and azimuth can be augmented with altitude and identification.

Transponders are transceivers, like DME. When the RADAR interrogation is received, the transponder transmits a specific response. For most GA aircraft, the response will be the identification (mode A), and the altitude (Mode C). The transponder concept comes from a technology developed around WWII, used for positively identifying friend or foe targets (IFF). Early targeting RADAR on aircraft could identify targets, but not who they were. Occasionally people will still call a transponder  an IFF box.

If the transponder is Mode S, the response will include quite a bit more information. The mode  S transponder has a payload capable of holding identification, altitude, and various other information, depending on mode. The message can be 56 or 112 (extended squitter or ES) bytes and include a 24 digit ICAO identifier assigned to each unique aircraft. Location and speed can also be encoded in the response. There are various modes the Mode S transponder will work in. (For a good article that covers much of the ModeS modes, see this EETimes article)

Usually aircraft transponders will transmit on 1090MHz. TCAS receivers and RADAR antennas will all be expecting to receive messages on 1090MHz. The transponder is mostly listening, but can be quite busy in class B airspace, with several TCAS units pinging traffic in the area. 

The code that is entered in the transponder is asigned by ATC. There are only 4096 unique codes, and some are reserved (IE 1200, 0000, etc). The numbers are limited to 0-7 or octal digits (octal = 8, and 0-7 are 8 distinct values). Octal is a throwback to early computers that were used for Air Traffic Control, and numbers were represented in octal values. On a busy day, there may be more than 4000 aircraft in the air at once, how does air traffic control keep conflicts out? Certain ranges of transponder (squawk) codes are reserved for local traffic (staying in the area, like training, or ferry flights). Other ranges are for long distance flying, some east, some west, depending on origin and destination. Occasionally, something unexpected happens, and two aircraft with the same transponder code appear in the same area, and the RADAR display will alert the controller to that.

The altitude that all the transponders send is pressure altitude. Pilots will set the altimeter on the ground to local barometric pressure. The altitude encoder attached to the transponder is not adjusted to local barometer. ATC will set their scope to the local pressure. Having ATC consistently reading the same uncompensated pressure will allow more consistent readings aircraft to aircraft. Sometimes pilots will forget to change their altimeter, or set it wrong, and this would cause trouble for ATC trying to figure out what everyone's altitude is. If ATC is saying the aircraft reporting altitude is significantly different than the pilot thinks they are flying, ATC may ask the pilot to stop reporting altitude, and the pilot will switch to Mode A.

Most Mode S transponders are capable or working in Mode A/C or just Mode A as well. Mode S transponders, with their large payloads can be used for ADS/B as well. ADS/B will require other  transponders in the area to be sending specific payloads, in order to plot the position on the receiving aircraft's display.

Transponders add a great deal to RADAR. Transponders will stay on aircraft even after the aircraft are switched to ADS/B. Eventually, the need for a transponder will be replaced by the ADS/B system, but that may be many years. 

Whats wrong with RADAR?

Ever see the news reports about the "World War II RADAR technology"? The headlines are usually provided by the FAA or other vendors when talking about NextGen technologies. RADAR has significantly improved since World War II. It is much more reliable, more consistent, and more accurate. The output now is mostly digital, and requires little adjusting.

All RADAR systems work by sending out a radio signal, and listening for that signal to bounce off a target, and timing the round trip of the signal. A passive RADAR signal is one where dish sends out a signal, and listens for the return. The passive RADAR message can only measure distance from the dish. Knowing the orientation of the antenna when the target distance was measured will allow the operator to know the range and azimuth of the target relative to the antenna.

The radio signal goes about the speed of light through the air, or about one foot per nanosecond, or about 5ms per mile, and remembering to double that for the round trip, will allow the RADAR system to determine the distance.

The RADAR dish is used to focus the transmitted signal, as well as the return signal. The pointy part near the bottom of the dish is the antenna for both the transmitter and receiver. The dish is a parabolic reflector, with the antenna at the focus point. The antennas are aimed at the dish. While the antenna does a good job of focusing the signal, it still goes out in a cone shape.

RADAR will detect various targets. The metal targets reflect the radio signals well. Other material will reflect at different levels. Most aircraft have metal somewhere, including tube and fabric, composite and wooden aircraft. Water also reflects radio signals. A large blob of moisture will show up on RADAR as a target. The processor on the RADAR unit will separate the blobs of moisture from the metal things. The blobs of moisture will be called weather, and the other metal objects will be considered primary targets.  

Many dishes have a secondary surveillance antenna on them as well. Secondary surveillance is used to listen for the transponder that is on many aircraft. The transponder on the aircraft will transmit the aircraft altitude, and some other data. The transponders will automatically transmit when they hear the RADAR interrogation signal.

Mostly there are two types of RADAR in use for civil aviation in the US, enroute and tracon. Enroute RADAR, or ARSR covers a radius of about 250 miles, and the dish rotates in about 12 seconds. Tracon RADAR covers about 60 miles, and the dish rotates in about 4.7 seconds. Both RADAR types can feed computers, that allow different people to see different views of the same data.

Since the RADAR signal go goes out in a cone shape, the exact position of the aircraft is less accurate the farther the target is from the RADAR antenna. The tracon RADAR will be more accurate than the ARSR RADAR since it is turning faster, and only is looking at shorter distances.

The RADAR signal can be blocked by buildings and terrain. Buildings and terrain can also reflect signals. Reflected signals can make the targets appear to be farther away. If an aircraft is opposite terrain relative to the antenna, it won't be picked up by the RADAR. Enroute charts will have a MSA altitude indicating the lowest altitude the RADAR can allow the controllers to see the aircraft.  

The RADAR units will output various channels, weather, secondary, and primary target data.  This data will be collected by computers, and be correlated to determine a track. Correlating the secondary target with the primary data will allow a track to know an aircraft speed, altitude and location. Correlating the signals will also need to remove bogus signals, like reflections, or smallish blobs of weather.

Newer technologies called multilateration is another way to find an aircraft. The multilateration will rely on the transponder on the aircraft. The ground station will have multiple receivers in known locations. A transmitter in the area will send out a signal, the transponder will detect the signal, and respond. The ground stations will measure the time it took to receive the signal,and the difference will tell the range and azimuth of the signal. The signal will contain the altitude.

Building RADAR sites can be expensive, building multilateration sites can be significantly less. If some acreage is available, the multilateral station can be a good choice to cover mountainous terrain, rather than building new RADAR sites in the mountains. The output of the multilateration system can feed the same computer systems that are used for RADAR displays.

Modern RADAR systems are quite flexible. Much different than the "World War II technology" the newscasters present. RADAR also has the advantage of not requiring any technology on the aircraft to work. Should an aircraft have a system failure, or an operator turn off a transponder. The FAA would like to decommission RADAR, but I believe long term, they won't completely. DOD and other organizations will require their existence. 

Is that helpful?

ASDI what is it?

When we go to FlightAware.com, FlightExplorer or any of the other flight tracking web sites or apps, they have a lot of good information. The information all comes from the FAA, for free! Ever think about the FAA and how they collect that information? How does is all come together? 

All over the country, there are RADAR sensors. These RADAR sensors are scanning the skies 24 hours a day, 7 days a week. The output of the RADAR sensor is sent to a computer, where the range and azimuth data is correlated to the transponder and altitude data. In the drawing below, the black line coming out of the RADAR dish is the interrogation signal, the black line coming back it the "skin paint" reflection signal, and the blue line is the transponder broadcast from the aircraft.

The computers correlate the transponder code to the flight plan, and take the RADAR returns and calculate a speed, altitude and track that the aircraft would be on. The flight plan helps determine where the airplane will be, based on the speed and time since last sample. 

All of the track information for all of the RADAR systems are sent to the FAA command center where they are made available for display in the various FAA systems that need the information (IE URET, TRACONs, ERAM, TFMA, etc).

One of those system that get the FAA RADAR data is the Aircraft Situation Display to Industry (ASDI). The ASDI data is availble to the airlines and other organizations in the aviation industry. The information includes flight plans, position reports, departure, and cancel flight plans. Using this information sites like FlightAware can present aircraft on maps for the general public. 

The ASDI data only contains non-blocked RADAR and flight plan data for aircraft with IFR flight plans in the US and some of Europe. There is an option available to private aircraft allowing them to block the ASDI data for competitive reasons (IE the president of ATT doesn't want the Verizon corporation to know about some special meetings with a partner or something).  

There are about 4 different kind of feeds of ASDI data. There is the internal FAA feed, the need to know real-time feed, the need to know real-time with European data, and a delayed feed. The real-time feeds are for the airlines and such to use for business reasons. The delayed feed is for the web sites visited by the general public. The delay is like 5 minutes, so it is good enough for people to know when to show up at the airport to get their loved one. 

There is a bit of information that can be derived from the ASDI data. Looking at the ASDI data, someone can determine which airports are taking delays with many aircraft holding. Other things can include looking at the projected track, and weather data to see when it might be best to re-route an aircraft because it is heading toward some convective activity. Airport operators can use the data to count operations relative to other airports, to help improve service. 

The ASDI data feed contains a lot of data. The position reports will be updates of aircraft positions every time the computers are aware of a position update (IE every 12 seconds for enroute RADAR, 5 seconds for TRACON, or 1 second for ADS-B). Typically Monday through Friday in the US there will be 3000-5000 aircraft in the air from 6am to 6pm. 

The FAA has many other services similar to ASDI. Most of it isn't as useful. ASDI is a great resource. 

What do you think.

ADS/B and RNP

There are many uses for the GPS data in and out of the aircraft. The GPS location is very accurate, normally. Knowing where the aircraft thinks it is can help ATC in many instances.

Shortcomings of RADAR

RADAR will send out a radio signal in a cone shape. The farther the aircraft is from the antenna, the larger the target will appear on the radar screen. The location displayed to the air traffic controller isn't as accurate when the aircraft is far from the antenna. The RADAR cannot "see" straight up either, so if the aircraft flies directly over the antenna, the software has to guess where the aircraft will be.

RADAR can only get range and azimuth information. The RADAR cannot determine altitude. Altitude information is sent from the aircraft in  the transponder message. When the aircraft transponder hears the RADAR interrogation, the transponder responds with the transponder code and altitude (with mode-s, there may be more information).

Most short range RADAR has about a 5 second sweep. The long range radars have a 12 second sweep. The sweep time is how long the RADAR antenna takes to turn once. The sweep time is how long it takes between aircraft updates. If the aircraft is going 600knots, and the sweep is 12 seconds, the aircraft moves about 2 miles between sweeps. 

The RADAR software has to do some correlations between the raw RADAR range and  azimuth information, and the transponder code altitude message. Usually, there is only one aircraft that comes into RADAR range at a time at the same point, so it is easy to correlate this, but occasionally two targets may appear at different altitudes at the same place. The software will occasionally get this wrong.

Why ADS/B is Better

ABS/B out messages from the aircraft will usually be the same quality. The GPS accuracy will be pretty consistent in an area, and be very accurate. The target drawn on the air traffic controllers screen will be the same size as the target moves across the screen.

The ADS/B message will contain both location and altitude information in a single message. The software will not need to correlate that data. During times when the ADS/B aircraft are operating in the RADAR environment, correlation will still be done with the raw RADAR and the transponder messages. The ADS/B information will only make correlation more accurate.

The ADS/B location is broadcast about once a second. The controllers screen will update every time it hears the ADS/B signal.

ADS/B will broadcast to the aircraft in the area without relying on ground station. The FAA having two frequencies, 978MHz and1090ES almost requires a ground station for ADS/B to work. There are other benefits to the ground stations, in that they will broadcast weather (FIS/B) and traffic (TIS/B) from non-participating aircraft.

RNP and ANP

In most of this and previous articles I have hesitated on specifying the accuracy of GPS location information. GPS accuracy changes during the day, and in certain locations. The current accuracy is defined as Actual Navigation Performance (ANP) and is measured in miles. Sometimes the ANP will be down to feet in all directions, sometimes it will be in miles.

To fly a GPS approach it is necessary to have an ANP of 0.3 miles or better. 0.3 miles means the receiver is able to pinpoint it's position to less than 1500 feet. The 0.3 mile value is called the Required Navigation Performance (RNP). To fly with any more precision, special training is needed. The FAA has many public RNP approaches with levels as low as 0.1 miles, or about 500ft.

This sounds pretty sloppy, 500ft is bigger than 10 houses.  Remember, the GPS receiver is calculating where it was when it heard the last update from the satellite, but the aircraft is carrying that receiver at 150-200kts on final.  The receiver is throwing values into the Kalman filter as fast as it can, and guessing that the pilot won't turn more that 3 degrees per second, assuming a mostly straight course.  It ain't easy, 500ft is pretty good.

Many commercial aircraft display the ANP and RNP values at the bottom of the navigation display. The display on a 737 is the instrument on the left (see slightly above this link).

NextGen future

There is a lot to NextGen technologies. Alaska has been playing with ADS/B since the late 1990's. The FAA is financing more Capstone work in Alaska even in the time of sequestering. Much of the enroute NextGen requires ERAM, but that is still in process, and is partially on hold until the FAA gets their financial situation in order.

The FAA is currently wanting to require all aircraft to participate in ADS/B out by 2020. There will be some challenges to that. With the current financial situation and things being on hold, will the FAA be ready for all aircraft to be using ADS/B? What about the Luscome 8F that never had an alternator, what will it use to power the GPS and transmitter needed? What will the FAA use to track aircraft with electrical failure?

Some airlines are equipping their aircraft with RNP and ADS/B. Some have had a challenge reaping the benefits from the upgrades. The other airlines are waiting until some indication they will reap some benefits. The FAA and the airlines are still trying to figure out the best time to move forward.