Enhanced Head-Up Display for General Aviation For the Quarterly Review of the NASA/FAA Joint University Program for Air Transportation Research Wednesday October 10th, 2001 Presented By: Douglas Burch Principal Investigator: Dr. Michael Braasch Douglas Burch Avionics Engineering Center Ohio University Athens, Ohio 45701 Phone: (704) 593-1526 Email: douglasburch@ieee.org Avionics Engineering Center Ohio University, Athens Project Sponsor: Joint University Program

Introduction General Aviation Instrumentation has undergone little change in the past 50 years. In 1999, 73% of the fatal accidents were caused by night Instrument Meteorological Conditions (IMC). IFR traffic is expected to increase by 2.5 percent per year over the next decade. Increase in IFR traffic might lead to a possible increase in GA accidents. In 1999, 75% of accidents caused by weather had fatalities. Another 73% of accident fatalities were caused by night Instrument Meteorological Conditions. The NALL Report lists several reasons for increased GA accidents; two that are applicable to this presentation are: Fewer cockpit resources-air carrier operations require at least two pilots; GA operations are predominately single pilot. Less weather tolerant aircraft, which generally must fly through the weather instead of over it. They also may not have the systems to avoid or cope with adverse conditions. The FAA is anticipating an average increase in IFR traffic of 2.5% per year over the next decade. With this type of increase in IFR activity one would expect an increase in accidents. An improved instrument display, like the eHUD, may help to avoid this.

Overview Motivation Behind eHUD Pseudo-Attitude Determination Current eHUD System Overview Flight Test Performance Requirement Analysis eHUD Architectural Overhaul Future Upgrades Motivation behind the eHUD project. Pseudo-Attitude Determination, what is really happening in the “little black box”. How the current eHUD is configured and some pointers covering what could be done better. The flight test and the data collected from these test. Architectural Overhaul, what is currently being done. Future upgrades to the system to make it comply with modern HUD configurations and in-house.

Motivation Behind eHUD Provide Visual Cues in IMC. Increase Situational Awareness in IMC. Reduce pilot training and recurrency requirements for flight in IMC. Keep the pilot looking out the window at the same time they are flying the instrument approach. Cost effective Head-Up Display. The main goal of the eHUD is to provide the pilot with visual cues that will increase their situational awareness during IMC flight conditions. The eHUD if implemented correctly should help reduce the recurrency requirements for flights in IMC. A human factors issue that needs to be dealt with is creating a display that keeps the pilot looking out the window during an instrument approach. For this type of Display to be used by the average GA enthusiast it must be cost effective.

Attitude The Merriam-Webster Dictionary defines attitude as the position of an aircraft or spacecraft determined by the relationship between its axes and a reference datum. Traditional Attitude: Three GPS Receivers, three Antennas. Expensive and Computationally Intensive. Pseudo-Attitude (Velocity Vector Based Attitude): Observable from a single GPS antenna. Cost effective to purchase and install. Attitude is the position of the aircraft in 3-dimensional space. This includes the roll, pitch and yawl which make up the Euler angles used for coordinate frame reference. Traditional GPS is the observed 3-space position of the aircraft typically calculated from the output of three GPS receivers. This is very expensive to implement and computationally intensive. Pseudo-Attitude is observable from a single GPS antenna by using the Velocity Vector to calculate the Flight Path Angle and the pseudo-roll angle. This is a very cost effective way to implement attitude determination in a GA aircraft but it does have its limitations.

Pseudo-Attitude Determination (Velocity Vector Based Attitude Determination) Developed at the Massachusetts Institute of Technology by: Dr. Richard P. Kornfeld Dr. R. John Hansman Dr. John J. Deyst The information on the following slides, regarding Velocity Based Attitude, was taken from “The Impact of GPS Velocity Based Flight Control on Flight Instrumentation Architecture” Report No. ICAT-99-5, June 1999. Self explanatory.

Reference Frame (North, East and the Local Vertical Down.) N × E = D Velocity Vector Vg = (VgN, VgE, VgD) When using the velocity vector to determine the pseudo-attitude a reference frame must be established. This is done by assigning the x-axis to the North and the y-axis to the East. By taking the cross product of the Northern and Eastern axis the vertical down is given. The Velocity Vector with respect to the ground is notated as Vg = (VgN,VgE,VgD). FNED: Earth-Fixed locally level coordinate system.

Pseudo-Attitude γ φ Xb Vg Yb Zb Flight Path Angle : γ Local Horizontal Reference Plane Yb Pseudo-Attitude is described by the Flight Path Angle Gamma and the Pseudo-Roll Angle Phi. The Body Frame is aligned with the x-axis going out the nose of the aircraft, the y-axis out the right wing of the aircraft and the z-axis pointing out the belly of the aircraft. The origin of the Body Frame is at the Center of Gravity of the aircraft. The Velocity Vector extends out from the C of G and is used to describe the motion of the aircraft as if it were a particle. The Reference Frame and the Body Frame are both instantaneously located at the C of G of the aircraft. The Body frame moves with respect to the aircraft attitude creating the Euler angles describing the aircraft’s pitch, roll, and yaw. Zb Flight Path Angle : γ Pseudo-Roll Angle : φ FB: Body-fixed orthogonal axes set which has its origin at the aircraft center of gravity.

Current eHUD Configuration GPS Antenna Novatel 10Hz Receiver GPS DATA…. Display Processor 600 MHz Laptop QNX OS The current eHUD configuration has a single GPS antenna located over the Center of Gravity for the Aircraft. A 10 Hz Novatel Receiver is used to provide the position and velocity data to a real time processor. The receiver sends the Position data at 5 Hz and the Velocity data at 10 Hz. The GPS seconds into the week of each string is used to match correlating position and velocity strings. This reduces the actual information rate of the receiver to 5 Hz for both the position and velocity information. The actual data rate between the receiver and real-time processor is 4800 baud. A 600 MHz Gateway laptop is used with QNX as the real-time operating system. The Flight Path Angle and Pseudo-Roll are calculated here and passed on to the Display Processor along with the Position information. The Display processor creates an image based on this and passes it to a flat screen CRT which is placed in the “dash” of the aircraft.

GPS Receiver Novatel GPS Receiver 10 Hz Velocity Data 5 Hz Position Data RS-232 serial port Novatel 10Hz Receiver Again, the GPS receiver provides Velocity data at 10 Hz and Position data at 5 Hz. This is used to determine the Pseudo-Attitude. GPS Receiver provides position and velocity information to the real-time processor for Pseudo-Attitude Determination.

Position and Velocity Strings $POSA,637,511251.00,51.11161847,-114.03922149,1072.436,… $SPHA,640,511251.00,0.438,325.034,2.141,… Position (POSA) GPS Sec into the Week. Latitude Longitude Height Velocity (SPHA) Horizontal Speed (m/s) Ground Track (degrees) Vertical Speed (m/s) Here is an example of the position and velocity strings that are being sent from the GPS receiver and parsed by the real-time processor. The velocity information is used in making the calculations for the Velocity Vector and ultimately the Pseudo-Attitude. VgN = Horizontal Speed * cosine(Ground Track) VgE = Horizontal Speed * sine(Ground Track) VgD = - Vertical Speed

Real Time Processor Gateway 600 MHz Laptop. QNX Real-Time OS GPS DATA…. Gateway 600 MHz Laptop. QNX Real-Time OS PCMCIA Card The real-time processor transforms the Velocity Data into the Velocity Vector, Vg = (VgN, VgE, VgD). This is used to calculate the Flight Path Angle and the Pseudo-Roll, which are sent to the display processor along with the position information. The Velocity Vector is created as follows using the right hand rule notation. VgN = Horizontal Speed * cosine(Ground Track) VgE = Horizontal Speed * sine(Ground Track) VgD = - Vertical Speed Once the Flight Path Angle and Pseudo-Roll Angle have been calculated, these are passed to the Display Processor along with the altitude, latitude and longitude.

DELPHINS Display Processor “Tunnel-In-The-Sky” Display Technology. Pioneered by Erik Theunissen at the Delft University of Technology, The Netherlands. Three-Dimensional representation of the outside world. This was designed for civil aviation to produce a tunnel-in-the-sky for a pilot to view through some type of display whether that be HUD or HDD. This is proprietary and we are unable to modify it, due to this it can take a while to have any custom implementations for he eHUD.

eHUD Display Display Image Flat Screen CRT As stated before, the display image is passed to a flat screen CRT which reflects the image onto the windshield of the aircraft. This needs to be modernized and will be discussed later in the future work section. Display Image Flat Screen CRT

Flight Test Scenario: Last Flight Test was June 8, 2001. Wind conditions were a concern. Consisted of two approaches on UNI runway 25. GPS Antenna mounted approximately above aircraft center of gravity. A total of three flight test were performed. The following data will show the results from the last of the three test. The wind was of concern during the flight test as is apparent in the following MatLab plots. The test consisted of two msised approaches on UNI runway 25.

Flight Path Here we see the flight path from the data collected from the current 10 Hz GPS receiver.

Altitude Profile Here is the Altitude profile from the same flight test.

Flight Path (Latitude, Longitude and Altitude with Respect to Mean Sea Level.) Here is the 3 dimensional flight path with respect to Mean Sea Level (MSL).

GPS Data Information at time t. Information Display at time (t – 4). Flight Test Results Results: Four-second delays noted in display image. Display seemed to indicate the correct aircraft attitude. Flyable once delay problem is sorted out. RCVR GPS DATA…. Display Processor 600 MHz Laptop QNX OS GPS Data Information at time t. Information Display at time (t – 4). The flight test showed that the we have a working system so to speak. There was a 4 second delay in the display. The aircraft would bank and then approximately four seconds later the Display would reflect this. Once this is worked out the system should be flyable.

Project Transition Have a system that demonstrates initial proof of concept. Have solid test data. Have a four-second delay problem. Conduct Performance Requirement Analysis. Check List: Novatel Receiver. Real time Processing of Velocity Vector. DELPHINS Display Processor and Imaging. When I took the project over 8 weeks ago I was given a system that demonstrated the basic or initial proof of concept for the eHUD. I have solid test data. The weak link here appears to be the real-time processing algorithm.

Problem Analysis Need: Increase in the number of GA accidents due to flying instrument approaches in IMC conditions. Goal: GA display that will help to mitigate problems associated with flying in IMC conditions by enhancing pilot’s situational awareness. Objective: - Must give proper representation of aircraft attitude. - Must be easy to interpret (Human Factors). - Must be cost effective (Single GPS unit). - Must keep pilot looking “out the window” (Human Factors). - Must be easy to mount in any GA aircraft.

Problem Analysis Continued Constraints: - Less than 500 ms delay for initial proof of concept. (Ultimately 100 ms delay or less.) - Image must be displayed with zero distortion. - Display size will be 22” by 22” for proof of concept. The constraints are very simple at this time. I have not yet had the opportunity to work with the full system only look at the flight data. These constraints are derived from the 4 second delay on the current configuration and the current display method.

eHUD Architectural Overhaul Phase 1: Update GPS Receiver to a Novatel OEM4 with 20 Hz position and velocity data (completed). Gather flight data with new receiver. Phase 2: Re-write Velocity Vector Attitude Determination code to insure timing issues are understood while processing the velocity vector. Phase 3: Find an alternative means to display attitude and position information to the pilot. I like to call my future work with the system an overhaul. I have a system, now I have to make it work and in real world conditions. This involves updating the GPS receiver, which I have already done. I now have 20 Hz position and velocity data. The improved software on the GPS OEM4 provides the velocity information directly from the Doppler effects of the carrier. Since the algorithm is the main suspect in the four second delay this will be re-written. Intensive Bench testing will then take place.

Novatel 20 Hz OEM4 Receiver Phase One Gather test data with new Novatel OEM4 20 Hz GPS Receiver. Verify flight data to insure a sufficient amount has been gathered before the winter arrives. Have two or three unique flight profiles to test against Real-Time Processor and what ever display option is used. Gather new test data. Verify. Would like to have multiple flight data. Novatel 20 Hz OEM4 Receiver

Phase Two Complete re-write of Velocity Vector Based Attitude Determination Algorithm. Sample new GPS data file at 20Hz to emulate real-time input from receiver against display option. GPS DATA…. Complete re-wriite.

Phase Three Bench test entire system with data file to insure that the correct display is being produced for a given flight profile. Data File GPS DATA…. Display Processor 600 MHz Laptop QNX OS Bench test to insure system is producing correct display on time.

Future Plans Implement phases one, two and three. Keep the development of the eHUD completely in-house. Use tools that will allow us to personally develop graphical displays, projection, etc. and not depend on others to make modifications. Update the Pilot Display to a modern implementation of Head-Up Displays. After insuring that the algorithm is working as expected and an intermediate display option is providing some type of attitude and position information to the user then a more permanent solution to the actual display processor will be found. Would like to not depend on outside proprietary software. Also the display processor is very large. There are alternatives to this that will be presented at the next JUP.

Modernization of Display Projector mounted to ceiling in cabin. Smoked glass allowing for “see-through” display. Typically, military aircraft have an infrared camera display in the cockpit providing flight information to the pilot. General Aviation HUD should follow this display convention. The main portion that needs to be modified is the pilot display.

Contact Information Research Associate: Douglas Burch douglasburch@ieee.org Principal Investigator: Dr. Michael Braasch mbraasch@oucsace.cs.ohiou.edu

References Kornfeld, R.P., Hansman, R.J., Deyst, J.J., The Impact of GPS Velocity Based Flight Control on Flight Instrumentation Architecture. MIT International Center for Air Transportation, Cambridge, MA. Report No. ICAT-99-5, June 1999. Eric Theunissen. Integrated Design of Man-Machine Interface for 4-D Navigation (1997) Delft University Press, Mekelweg 4 2628 CD Delft, The Eric’s Web page: www.tunnel-in-the-sky.tudelft.nl.