D 7 - TRIMIS



DenseTraffic

IST-2000-29638

Deliverable 7.8

Final Report

| |

|Contract Start Date: |June 2001 |Duration: |30 Months |

|Project Technical Coordinator: |Dr. Claudio Hartzstein |RoadEye FLR GP |

|Partners: | | |

| |Groeneveld Groep B.V |The Netherlands |

| |United Monolithic Semiconductors S.A.S |France |

| |EADS Deutschland GmbH Microwave Factory |Germany |

| |RoadEye FLR GP |Israel |

| |ERA Technology Ltd. |United Kingdom |

| |DAF Trucks N.V. |The Netherlands |

| | | | |

|Document No. |9638E-SYS-F-D-Rep-03 |Rev |03 |Project |Dense Traffic |

|Page 1 of 46 | | | |Class |Public |

| |Name |Date |Signature |

|Prepared: |Dr. Claudio Hartzstein |Aug-04 | |

|Revised: |Ilan Bitton |Aug-04 | |

| | | | |

|Released: |Dr. Claudio Hartzstein |Aug-04 | |

|[pic] |Project funded by the European Community under the “Information Society |

| |Technology” Programme (1998-2002) |

REVISION HISTORY

|Revision |ECN |Description |Date |

|01 | |Preliminary Draft |Jan-04 |

|02 | |Including comments from reviewers |Mar-04 |

|03 | |+ tests results and roadmap of development and market |Dec-04 |

| | |development | |

| | | | |

| | | | |

| | | | |

| | | | |

| | | | |

| | | | |

| | | | |

Table of Contents

1 Project Overview 4

2 Project Objectives and Approach 6

3 Project results and achievements 6

3.1 System Description 7

3.2 Antenna 8

3.3 RF Module 15

3.4 Electronics 16

3.5 Signal Processing 17

3.6 Tests Results 24

3.7 Achievements and Risks 24

3.8 Contribution to Programme/Key Action Objectives 25

4 Community Added Value and Contribution to EU Policies 26

4.1 European dimension 27

4.2 European added-value 28

4.3 EU policy 28

4.4 Contribution to Community Social Objectives 29

4.4.1 Quality of life 29

4.4.2 Employment 30

4.4.3 Conservation 30

4.5 Economic Development and S&T Prospects 30

4.5.1 Exploitation 30

4.5.2 Strategic Impact 31

4.5.3 Dissemination 32

5 Workplan 34

5.1 General Description and Main Deliverables 34

6 Description of Dissemination Plan 38

6.1 Conferences/seminar/workshop 39

6.2 Publications 40

6.3 Web presence 40

6.4 Standardisation : 40

6.5 Other : 40

7 Project management and co-ordination aspect 41

8 Outlook 41

9 Conclusions 42

9.1 Radome 42

9.2 Antenna and Manifold 42

9.3 RF Transceiver 42

9.4 Electronics 43

9.5 Power Supply 43

9.6 Housing 43

9.7 Software 43

9.8 Algorithms 44

9.9 Development Roadmap and Market Penetration Strategy 44

9.10 Conclusion of the Conclusions 46

Project Overview

Present Forward Looking Radar (FLR) sensors commercialized in several high end luxurious cars, which can be classified as “First Generation” sensors, opened the market of the Adaptive Cruise Control to the general public. These systems have reached an adequate level of maturity and performance that allows the customer to rely on the ACC functionality and feel comfortable. Still, these systems suffer from several shortcomings: a) limited performance in azimuth angle coverage and short distance detection, and b) high cost. The very low market penetration of the ACC compared with the predictions made several years before, can be attributed mainly to the high production cost which when multiplied by the usual factors of the automotive industry, brings the price to the final customer in the 1500 to 2500 Euro bracket. The lack of understanding from the general public of the advantages of an ACC system together with their high price (compared with other products like car stereo, navigation systems, etc…) has, to our understanding, limited the market penetration to a few 10 kunits/year instead of the expected 1Munits/year or more.

The “DenseTraffic” project addresses to these two shortcomings. Its primary objective is to develop and demonstrate a Forward Looking Radar Sensor (FLRS) with improved capabilities that will allow operation in Stop&Go modes and early detection of Cut-In situations. This will enhance the functionality of the sensor in an Adaptive Cruise Control system. Additionally, but no less important, is to demonstrate the feasibility of a low-cost, high-volume production design that will allow the product to be mass produced.

The low cost objective was achieved with a single box design utilizing low cost Thyxomolded Magnesium technology for the antenna and housing and Monolithic Millimeter Wave Integrated Circuit (MMIC) technology for the RF transceiver. The performance was improved by using a multi-beam antenna design that enables wider azimuth coverage and the use of wide bandwidth waveforms digitally modulated by means of a Phase Locked Loop driven by a Direct Digital Synthesizer. We developed a flexible, adaptive waveform generation logic that optimizes the FLR performance in every road scenario. The waveforms and the corresponding signal and information processing allow a seamless transition between normal ACC and Stop&Go (in highways) regimes.

Intensive tests in Israel and Europe (including Lapland) in all kind of weather conditions and road scenarios, have demonstrated the superior performance of this FLR and a sound Bill of Materials shows that the production cost can be lowered to a level compatible with OEM’s price requirements.

This advanced driver assistance system will improve safety in dense traffic and reaction to emergency situations by providing enhanced performance at a lower cost, allowing a greater market penetration.

The “DenseTraffic” project was performed by a consortium constituted by the following partners:

|DenseTraffic Partners |

|Groeneveld Groep B.V (Coordinator) |The Netherlands |

|United Monolithic Semiconductors S.A.S |France |

|EADS Deutschland GmbH Microwave Factory |Germany |

|RoadEye FLR GP (Scientific Coordinator) |Israel |

|ERA Technology Ltd. |United Kingdom |

|DAF Trucks N.V. |The Netherlands |

The roles of the partners were as follow:

“Groeneveld Groep” was the Financial Coordinator and responsible for the contractual management of the consortium and the responsible for the distribution of the funds received from the EC and the financial reporting to the EC. As major shareholder of RoadEye it hosted RoadEye’s activities in Europe (data collection, integration and testing with OEM’s and Tier1 suppliers, and dissemination activities).

“UMS” developed the MMIC’s used to build the RF module of the FLR sensor. It also

“EADS” developed the multi channel RF module based on the MMIC’s supplied by UMS.

“RoadEye” was responsible for the technical coordination of the whole project. It performed the system engineering and design of the radar, produced the specifications for the deliverables of the rest of the consortium partners. RE developed the electronics, housing, algorithms and integrated and tested the radar.

“ERA” developed the multi beam antenna and together with RE participated in its industrialization.

“DAF” participation was reduced to nil during the course of the project since the integration and testing of the radar was performed in passenger cars which were much more available and easier to operate than trucks while from the point of view of the radar host, there is no difference between a passenger car and a truck. The tasks from DAF were transferred to TNO as a subcontractor of Groeneveld.

The DenseTraffic “Final Report” document contains a system description, a description of the most relevant benefits and important risks of the DenseTraffic application.

Project Objectives and Approach

The project's primary objective is to develop and demonstrate a Forward Looking Radar Sensor (FLRS) with improved capabilities that will allow operation in Stop&Go modes and early detection of Cut-In situations. This will enhance the functionality of the sensor in an Autonomous Cruise Control system. Additionally, but no less important, is to demonstrate the feasibility of a low-cost, high-volume production design that will allow the product to be mass produced. These objectives will be achieved with a multi-beam antenna utilizing metallised molded plastic and a multi-channel RF transceiver using MMIC technology. The FLRS will consist of a single, multi-beam, integrated sensor and include unique built-in sensor self-test capability and algorithms for adaptive waveform generation and multiple target tracking. This advanced driver assistance system will improve safety in dense traffic and reaction to emergency situations by providing enhanced range resolution and angular coverage.

Project results and achievements

The DenseTraffic project has been conducted as follows:

The main system components, including multi-beam antenna, MMIC chipset, RF transceiver module, and electronic hardware and FLRS algorithms were developed based on specifications developed during a prior system engineering phase. The FLR sensor hardware integration and development of embedded software for radar signal processing was performed and tested using advanced tools for the collection and analysis of data in test vehicles. These vehicles (in open and close loop) have been used to evaluate and validate the system in high and low speed, Stop&Go and Cut-In situations, including short- and long-term validation and performance analyses.

In the following subchapters we shall describe in some detail the work performed in the different work packages of DenseTraffic.

1 System Description

The RoadEye FLR sensor principal characteristics are summarized in the following table:

|RoadEye FLR principal characteristics |

|Frequency band |76 – 77 GHz |

|Transmitted bandwidth |variable, up to 400 MHz |

|Transmitted waveforms |FMCW and Range Doppler |

|Waveform generation |PLL + DDS |

|Separate single Transmit and multiple Receive beams |

| | |Az |El |

| |Transmit beamwidth |12 deg |4 deg |

| |Receive beamwidths | | |

| | |L, C and R |4 deg |4 deg |

| | |LL and RR |12 deg |4 deg |

| | |GL and GR |~ 20 deg |~ 10 deg |

|Molded magnesium antenna and housing |

|Glass filled Ultem radome with polarization grid in the internal surface. Heating element for ice and snow |

|removal. |

|Single RF module with one Tx channel and 8 Rx channels |

|8 parallel baseband amplifying channels with automatic gain control |

|Parallel sampling and processing of 8 analog channels |

|Flexible FMCW or Range Doppler processing |

|Multiple target tracking and parameter extraction |

|CAN bus interface for ACC implementation |

|Fast communication channel for raw data collection |

|12V or 24V power supply with temperature shut down |

|MEMS rate gyro in horizontal separate card |

| | | | | |

| | | | | |

2 Antenna

The basic principle of operation of RE’s FLR is better understood by looking into the patterns of the Transmit and the multiple Receive beams:

The Transmit pattern acts as an illuminator without angular discrimination capability. It has a strong intensity in the central 12 degrees coverage and it has intentionally increased sidelobes to broaden the illumination to wider angles but at reduced ranges.

The receive patterns shown in Fig. 2 implement the azimuth discrimination on receive. The three beams: Left, Center and Right (L,C and R) are dedicated to the detection of targets at long distances in a narrow angular coverage of about 12 degrees. The broader beams: Left-Left and Right-Right (LL and RR) are used for the detection of nearer targets at broader angles. The last two beams: Guard-Left and Guard-Right (GL and GR) are used as sidelobe cancellers.

The resulting two way patterns are shown in Fig. 3. and Fig. 4:

The angular coverage limited by the Guard antennas is of about (25 deg, but energy considerations limit the actual coverage to (20 deg depending on the target radar cross section. In Fig. 4 we show the realization of a rectangular coverage utilizing variable beamwidth beams.

This wide azimuth field of view pattern allows the FLR to detect and track targets both at long distances in straight segments of the road and at shorter distances in sharp curves. It is also this pattern that allows the early detection of “cut in” situations, which addresses to one of the most important customer complaints.

The patterns shown in Fig. 1 to 4 have been implemented in an antenna assembly described in Fig.5. The oval aperture at the center corresponds to the Tx beam. The circular aperture to the right includes the R, C and L beams. The oval aperture to the left includes the LL and RR beams. The two small horns at the extreme left and right are the GL and GR beams.

Fig. 6 shows a photograph of one of the five machined prototypes delivered by ERA. These very expensive prototypes were used for the validation of the antenna design and the integration of the first prototypes of the FLR.

The next step was the development of a mass production technology for the antenna. Our first choice for the antenna fabrication was metallized plastic (Ultem) as it was stated in the original proposal, but it was soon obvious that thermal considerations (the need of a heat sink for the heat developed in the RF module and conduct it to the ambient) required a good thermal conductor: i.e. a metal. Unfortunately, aluminum casting could not deliver the high tolerances required for a 77 GHz antenna without additional tooling. This solution was too expensive for a mass produced FLR.

After a thorough search and analysis, the thyxomolded magnesium technology was chosen. This moulding technology is very similar to plastic molding, in which the magnesium is at a low temperature where it is partly solid and partly liquid. The mechanical tolerances of the resulting piece are similar to the obtained with plastic without the need of additional tooling.

The design made by ERA was transformed by the RoadEye team of mechanical engineers into a moldable design (draft angles, rounded corners…). The moulds were machined and after two rounds of corrections, the final magnesium parts were fabricated. In Fig. 7 the magnesium parts that compose the antenna assembly are shown. In the top row at the left, the lower half of the manifold and the front housing is shown. In the center the upper half of the manifold with the feeds and to the left the reflectors are shown. At the bottom of the picture we see the radome with the printed horizontal conductive lines that function as a polarization filter.

Several tenths of molded magnesium antennas were produced. Part of them were tested and compared with the original aluminum to validate the new design. Others were used to develop a coating that protects the antennas from corrosion and subjected to a battery of environmental test (salt spray, temperature cycling…) until they met a basic set of tests.

The performance of the antenna is shown in Fig. 8. The measurements were made in an anechoic chamber fully equipped for 77 GHz measurements. The graph shows a high accuracy measurement region of about (10 deg. Beyond this region, the target is only detected in a single beam (LL or RR) and the angular measurement is therefore coarser but sufficient for continuous tracking of close range targets. The shown accuracy has been obtained with an algorithm locally optimized. Tests over wide temperature range and waveforms with a global optimization of the algorithm show some degradation of the angular accuracy but still in the (0.25 deg bracket.

3 RF Module

The multi-beam antenna concept of RoadEye’s FLR is supported by a multiplicity of parallel receive channels. The RF transceiver circuit was developed by EADS utilizing existing MMIC’s from UMS with additional two MMIC’s especially designed and produced by UMS required by RE’s design.

The schema of the RF circuit is depicted in Fig. 9.

A wide band VCO produces a signal that after being multiplied three and two times is transferred to the Tx antenna.

A sample of the transmitted signal is derived to a network of divide and amplify elements whose output drives four double mixers. Before each mixer, an additional Low Noise Amplifier was added to improve the Signal to Noise Ratio.

An additional narrow band fixed frequency DRO oscillator is used to downconvert an intermediate step of the transmitted signal and this low frequency signal is utilized to close a Phase Locked Loop which drives the VCO of the transmit chain.

Twenty prototypes of this RF module were produced and utilized to build FLR prototypes. These modules have excellent performance but they suffer from an important drawback: they are extremely expensive. Several contributions account for the RF module high cost:

– Very large Gallium Arsenide area.

– Non pasivated MMIC’s which leads to the high hermeticity requirement (fine leak level)

– Large total size of the Module (60 mm ( 60 mm)

– Low yield due to the previous reasons

It is obvious that the low cost goal of this project could not be met by the RF module developed by EADS, and therefore a Cost Reduction Program was initiated with UMS in which a bold step that will substantially reduce the cost of the RF components of the FLR is considered. The general concept is the following:

– Sharply reduce the Gallium Arsenide area by deleting the LNA’s and the distribution network of the local oscillator at 77 GHz

– Passivate the MMIC’s to relief the hermeticity requirement from the package

– Move to a distributed architecture which reduces the size of the packages and increases the yield

– Search for low cost packaging technologies: i.e. ceramics or plastic.

Practical decisions towards these goals have already been taken and work is in progress.

4 Electronics

The supporting electronics include:

– Digitally controlled DDS based Synthesizer and Waveform Generator.

– 8 parallel IF amplification and filtering channels with Automatic Gain Control capability

– 1 MHz / 12 bit parallel sampling of all channels

– DSP and CAN controller

– 12V / 24V power supply for car and truck applications

– Radome heater circuitry with temperature control

– Parallel fast communication channel for raw data collection

– CAN and Power connectors

The first electronic suit of cards was partly developed by EADS and RoadEye and the real estate was assigned in advance, based on estimations. After producing several prototypes with 5 cards, a size reduction program was initiated in which the number of cards was reduced to 3. Fig. 10 shows the present electronic configuration. In this configuration provision has been made to include a MEMS rate gyro.

From right to left, we see in Fig. 10

a) Ultem radome with heating element. The heating element performs two tasks simultaneously: Heating the radome if the external temperature drops below some threshold, and acting as polarization filter as an integral part of the antenna.

b) Front housing and back of the manifold. The RF module is in contact with the smooth surface. The 9 (1 TX and 8 RX) waveguide interfaces are clearly seen and around them a quantity of holes for the screws that hold the RF module into position.

c) RF module with two small cards around it with the PLL and IF analog circuitry.

d) DSP, FPGA and DDS card

e) Power supply card inserted in the back housing. Adjacent to the left wall of the housing, the housing of the gyro is apparent.

5 Signal Processing

A very large amount of effort has been invested in the development of a suite of algorithms which one hand are sophisticated enough to cope with the very complex situations observed in automotive scenarios, both in highways and in city, and on the other hand they are as simple as possible that can be developed, tested and proved reliable.

Simple algorithms can be developed based on models and simulations, but they will not be adequate in real situations. On the other hand it is impossible to predict and simulate the variety of situations in the real world. Therefore, the algorithm development at RoadEye has been performed, as it is usually done in the Radar community, through a constant cycle of data collection, algorithm development, testing on the data base, algorithm refinement, additional data collection, and so on.

The data collection setup basic requirement was the ability to collect the raw samples directly from the ADC (Analog to Digital Converters) for continuous periods of time of about 2 minutes together with a synchronized video picture of the road scenario. Several hundreds of Gigabyte of data was collected this way in Israel and several countries in Europe (Holland, Germany, Sweden…) in various traffic scenarios (fast highways, country roads, dense city and climatic conditions (hot, cold, very cold, light and strong rain, snow and strong snow, fog…). The data collected was essential to the algorithm development. It is impossible to imagine the development of a radar sensor for ACC application without this tool.

After some two years or more of algorithm development, the level of false alarm became low enough that the original method of just recording and catching random events was not effective. We were generating too much no longer relevant data. We therefore developed an application that keeps a circular buffer of data and when some specific interesting situation happens, then we press a button and download to the hard disk some ten seconds before the specific event and record until stopped. This procedure allows the collection of data relevant to the further improvement of the algorithms.

The signal processing implemented is best described in the flow diagram shown in Fig. 11. The important difference with other existing FLR’s is that the transmitted waveform can be modulated at will with simple software commands. This freedom entails the development of a Waveform Selection Logic that should determine which of the waveforms will be transmitted and processed depending on the situation.

At the beginning of the development, we relied mainly on FMCW waveform transmition and processing. The FMCW principle of operation is shown in Fig. 12. We obtained rather good performance in Israel and in Holland, but after some data collection campaigns in Germany, we observed that the very strong returns from the poles that hold the guard rails hindered the target detection and distorted the angular measurement. In Fig. 13 a graphical explanation of this phenomenon which results from the lack of discrimination between moving targets and static objects.

We therefore decided to implement a Range-Doppler waveform to overcome this difficulty. The principle of operation of the Range-Doppler waveform is described in Fig. 14. This waveform has the Doppler discrimination capability (as shown in Fig. 15) and is widely used in defense radar systems but it is the first time that has been applied to FLR’s in automotive applications.

The implementation of the Range-Doppler technique included the design of several different waveforms with range and range resolutions appropriate to the varying scenarios. Each waveform was implemented with three different PRI’s (pulse repetition intervals) to solve the velocity ambiguity.

The performance of the FLR has been extended to very short distances and it has been tested in Stop&Go mode. The present design of the RF module has too much gain and the receiver saturates when the target is too close causing sometimes the loss of the target at close distances ( ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download