Good MMIC test data correlation between wafer tests and substrate tests was observed. Test data correlation between the supplier and Delphi was also established. Initial results indicate excellent performance from the VCO and 76 GHz amplifiers. Existing chips will be useable in 1st iteration transceivers. The frequency doubler has high VSWR and temperature issues and the VCO has yield issues due to temperature variation. A design review was held with the supplier to address the VSWR and temperature issues. Actuator control software was found inadequate for dynamic tests due to mechanical resonance leading to redesign of the antenna actuator system. Sidelobe levels on the EDU (Engineering Development Unit) antennas were found to be inadequate for good on-road performance. EDU antennas will be replaced with prototype units which perform as specified.

Plans through December 2000

During the next reporting period, the EDU transceiver design will be completed, and the second iteration chip set will be released to fabrication. Initial sensors will be replaced with upgraded versions for on-road test.

Figure 4.1 Task B1A Schedule

4.2 Auto Alignment Algorithm Development (Task B1B)

Implement an algorithm that electronically adjusts the sensor for mechanical misalignment due to vehicle wear and tire alignment.

Automatic alignment development starts with a generic study of possible system approaches that include use of radar data, external sensor data, and external vehicle data. This study is followed by selection of one or two technically feasible approaches as possible solutions. Detailed development of the selected approaches will then take place, followed by design, fabrication and bench testing of the completed approaches into the CW sensor and road evaluation of the integrated units. The final version will be used in the deployment vehicles.

Initial algorithm development has been completed and the algorithm has been tested. The following activities have been completed:

1. Define objectives and requirements
2. Perform algorithm development
3. Perform software development
4. Perform bench and road test and evaluation.

Research Findings

The basic algorithm is functioning quite well with the alignment process found effective over "long" periods of time. Azimuth offset angles were correctly characterized and removed by the alignment process. Occasional peak errors of short time duration were observed in the data and found to be caused by curves. The intent to warn the driver if the data was not converged has not been possible due to these curve induced data errors. Field-testing is continuing to isolate the effect that caused temporary peak errors in the correction process. The remaining task is to identify and correct causes of the curve induced errors resulting in separating true misalignments from curve events, and the ability to caution the driver that the unit is re-aligning.

Data gathering will continue as will development of alternate approaches to address the problem with curves. Road test data gathering has slowed, though the basic algorithm is functioning well in a variety of test vehicles. This task has fallen behind schedule with regards to data collection and algorithm enhancement; A recovery plan is being executed.

Figure 4.2 Task B1B Schedule

Objectives

Implement an algorithm that detects and warns the driver when the sensor is blocked by dirt, slush, or other material.

Approach

Several possible system approaches are being investigated including the use of radar data, external sensor data, and external vehicle data. This study will be followed by selection of two to three technically feasible approaches as possible solutions to the radome blockage problem. The best approach will be selected based on bench testing, simulation, and road evaluation and will be implemented in sensors used for deployment vehicles.

Work Accomplished

Two generations of radome blockage detection algorithms were designed and implemented. The upgraded technique uses non-coherent integration of main beam clutter. An alternative technique utilizing radar track data was also defined. A data collection test plan was created to evaluate more comprehensively the candidate techniques. Initial data collection was completed for the initial, upgraded, and alternate techniques. Data reduction is in progress. The following activities have been completed:

1. Define objectives and requirements
2. Perform algorithm development
3. Perform software development
4. Perform bench and road test and evaluation.

Research Findings

This task has fallen behind plan and a recovery plan is being worked. No problems have been identified to date. Some conceptual concerns with radome blockage detection have been identified regarding detection reliability. Concerns over detection reliability in a partial blockage condition or during heavy snowfall or when the host vehicle is parked will be investigated, and adequate test and validation scenarios are being defined.

Plans through December 2000

Data collection and analysis will be completed. Algorithm enhancements will be implemented and the schedule recovery plan executed.

Figure 4.3 Task B1C Schedule

Objectives

Implement an algorithm that classifies bridges as "safe" overhead obstacles and does not slow the vehicle unnecessarily.

Approach

Several possible system approaches will be investigated using radar data from the wide field of view multi-beam ACAS/FOT radar sensor. This study will be followed by selection of two to three technically feasible approaches as possible solutions to the bridge rejection problem. The best approach will be selected based on bench testing, simulation, and road evaluation and will be implemented in sensors used for deployment vehicles.

Work Accomplished

Several candidate approaches for bridge discrimination were investigated. Based on initial data collection and analysis, one approach was selected as the initial design. Initial road test and evaluation was completed for the initial design. A second design iteration was completed after analysis and review of first iteration design test results. Road test data was collected to evaluate the second iteration design. Data was collected on approximately 150 bridges and overhead signs in the Detroit and Indianapolis metropolitan areas. Tests were performed against 23 stopped vehicle situations involving 15 different vehicles. Initial data reduction and evaluation was completed. Definition of further algorithm improvements is in progress. The following activities have been completed:

1. Define objectives and requirements
2. Perform algorithm development
3. Perform software development
4. Perform bench and road test and evaluation.

Research Findings

It was determined that the multipath lobing structure is not a reliable discriminant because multipath lobing on some stopped vehicles can lead to significant amplitude deviation. Two amplitude discriminants (deviation and slope) were also investigated for potential use in distinguishing bridges from valid in-path stopped objects. It was found that amplitude deviation is not a reliable discriminant. Amplitude slope is a reliable discriminant, but this method takes time (range closure) to develop, which can lead to delayed recognition of valid in-path stopped vehicles.

Plans through December 2000

The second iteration algorithm design and implementation will be completed. Bench testing of the second iteration design will be completed and road test and evaluation of this design will be started.

Figure 4.4 Task B1D Schedule



[TITLE PAGE]      [TABLE OF CONTENTS]
[1 Executive Summary]     [2 Introduction]     [3 System Integration]     [4 Forward Radar Sensor]
[5 Forward Vision Sensor]     [6 Brake Control System]     [7 Throttle Control System]
[8 Driver-Vehicle Interface]      [9 Data Fusion]     [10 Tracking & Identification]     [11 CW Function]   
 [12 ACC Function]     [13 Fleet Vehicle Build]      [14 Field Operational Test]
[Appendix A]     [Acronyms]