Periodic Reporting for period 2 - PYCSEL (PYroelectric Conformable SEnsor matrix for Large area applications in security and safety)
Berichtszeitraum: 2018-07-01 bis 2019-12-31
CONTEXT
Biometrics sensors for personal identification undergo a huge evolution since a few years, as it is not only reduced to governmental and enterprise markets anymore. Nowadays, fingerprint sensors can be found in consumer electronics such in smartphones and laptops. Basically, two main technologies can be mostly found in fingerprint sensors: one based on optical devices and another based on Silicon devices. The first one is a robust and mature technology which enables a high acquisition/ high resolution surface at a reasonable price, and allows to capture multiple fingerprints concurrently. However, prism-based optical sensors demonstrate some limitations: they are heavy, thick, bulky, non-conformable and raise concerns regarding integration, product design, portable uses and cost. On the other hand, Silicon devices, i.e. mostly capacitive, pressure or thermal devices, are used for consumer electronics, due to their small size and easiness for integration. However, when large sensing areas are required, it is not a possible solution for cost reasons.
CONCEPT & OBJECTIVES
PYCSEL will address the bottlenecks of current fingerprint identification systems using flexible electronics (also called ‘Thin and Organic Large Area Electronics’ (TOLAE)) technology. These thin conformable large-area sensors present the advantage:
- To be free of form factor constraints during integration;
- To show enhanced ergonomics in future innovative cost-efficient fingerprint identification systems while enabling large area/high resolution acquisition.
For this purpose, PYCSEL project will develop a low cost thin and large area fingerprint sensing surface enabling the personal identification via the development of a TOLAE technology, combining an organic sensor with a TFT matrix on a plastic foil. Based on the fact that personal recognition requires high resolution (500 dpi) and large (1 up to 4 fingers) sensors, the project focuses on the design, development and integration of a printed pyroelectric PVDF-based sensor layer on a IGZO TFT active matrix on foil and connected to an electronic driver and readout board, resulting in a thin fingerprint conformable sensor with no need for any optical bulky and/or costly extra components integration. Multiple fingerprints capture will be possible with the resulting large area hybrid system whose conformability allow easy further integration and ergonomic use especially for high growth and high value portable security uses.
Among other things, it will require developing a manufacturable process for a 500 ppi TFT backplane and a reliable fingerprint pyroelectric sensor, compliant with the 500 ppi high resolution requirements, including encapsulation (mechanical protection and shielding) and specific sensor poling methods. A specific demoboard will be also developed for each of the two formats of sensor (the first one is 256x256 pixels corresponding to a '1 finger' imager, and the second one is 1600x1500 pixels enabling a '4fingers' image capture).
The different steps of the PYCSEL project are depicted in the figure 1 below.
Biometrics sensors for personal identification undergo a huge evolution since a few years, as it is not only reduced to governmental and enterprise markets anymore. Nowadays, fingerprint sensors can be found in consumer electronics such in smartphones and laptops. Basically, two main technologies can be mostly found in fingerprint sensors: one based on optical devices and another based on Silicon devices. The first one is a robust and mature technology which enables a high acquisition/ high resolution surface at a reasonable price, and allows to capture multiple fingerprints concurrently. However, prism-based optical sensors demonstrate some limitations: they are heavy, thick, bulky, non-conformable and raise concerns regarding integration, product design, portable uses and cost. On the other hand, Silicon devices, i.e. mostly capacitive, pressure or thermal devices, are used for consumer electronics, due to their small size and easiness for integration. However, when large sensing areas are required, it is not a possible solution for cost reasons.
CONCEPT & OBJECTIVES
PYCSEL will address the bottlenecks of current fingerprint identification systems using flexible electronics (also called ‘Thin and Organic Large Area Electronics’ (TOLAE)) technology. These thin conformable large-area sensors present the advantage:
- To be free of form factor constraints during integration;
- To show enhanced ergonomics in future innovative cost-efficient fingerprint identification systems while enabling large area/high resolution acquisition.
For this purpose, PYCSEL project will develop a low cost thin and large area fingerprint sensing surface enabling the personal identification via the development of a TOLAE technology, combining an organic sensor with a TFT matrix on a plastic foil. Based on the fact that personal recognition requires high resolution (500 dpi) and large (1 up to 4 fingers) sensors, the project focuses on the design, development and integration of a printed pyroelectric PVDF-based sensor layer on a IGZO TFT active matrix on foil and connected to an electronic driver and readout board, resulting in a thin fingerprint conformable sensor with no need for any optical bulky and/or costly extra components integration. Multiple fingerprints capture will be possible with the resulting large area hybrid system whose conformability allow easy further integration and ergonomic use especially for high growth and high value portable security uses.
Among other things, it will require developing a manufacturable process for a 500 ppi TFT backplane and a reliable fingerprint pyroelectric sensor, compliant with the 500 ppi high resolution requirements, including encapsulation (mechanical protection and shielding) and specific sensor poling methods. A specific demoboard will be also developed for each of the two formats of sensor (the first one is 256x256 pixels corresponding to a '1 finger' imager, and the second one is 1600x1500 pixels enabling a '4fingers' image capture).
The different steps of the PYCSEL project are depicted in the figure 1 below.
Main activities from M1 to M18:
First activity of this period was the definition of the sub-system specifications requirements for the PYCSEL fingerprint sensor prototypes to be developed for governmental and automotive use cases. In parallel, electrical and thermal modeling actions have been processed in order to help to define the major parameters (timings, signal level) as well as to orientate the stack definition of the PYCSEL sensors.
From a process point of view, the compatibility study of the pyroelectric frontplane with the TFT backplane has been achieved leading to significant process development and adaptation in the both TFT and pyroelectric components process flows.
PYCSEL 'one-finger' sensors design represented also extensive works during this period and resulted in providing all the necessary mask set tapes for the manufacturing of the first fingerprints sensors (first sensors are related to one finger capture; in the next period new design will be done related to a ‘4-finger’ capture imager).
The system integration, related to the development and manufacturing of the driving and readout electronics, has been also addressed during this first period and first monitoring tests of the sensors have been achieved leading to demonstrate the pyroelectric signal detection.
Main activities from M18 to M36:
- the driving and readout electronics of the 'one-finger' (256 x 256 pixels) sensor was finalized enabling the first acquisition demos and campaign.
- the specific design of the '4-finger' capture imager has been achieved taking into account the lessons learned from the 'one-finger' sensor realised during the first period. We created four different timing options for the readout limiting the number of I/Os and accelerating the acquisition,
- IGZO backplanes for the large area fingerprint sensors were next manufactured followed by the printing process of the large area pyroelectric sensor layers,
- the design and development of a new demoboard enabling the acquisition of 1500 x 1600 pixels has been necessary with specific throughput challenges solved by using USB3 interface and specific memory buffer,
- in the meantime, the quality of the 'one-finger' capture imager has been assessed and compared to a commercial optical imager, with good results. A specific filter for thermal images has been developed to fit the recognition algorithms. An acquisition campaign was performed and enabled to evaluate the recognition performance of the 'one-finger' sensor showing good results regarding the TRL level of the PYCSEL sensor. We also showed the low impact of a slight curvature of the fingerprint sensor on the image,
- Last, the PYCSEL consortium was awarded with the "best publicly funded project demonstrator" at the LOPEC OE-A competition 2019. A paper was also presented at the IEDM 2019 and an other one has been accepted at the SID 2020.
First activity of this period was the definition of the sub-system specifications requirements for the PYCSEL fingerprint sensor prototypes to be developed for governmental and automotive use cases. In parallel, electrical and thermal modeling actions have been processed in order to help to define the major parameters (timings, signal level) as well as to orientate the stack definition of the PYCSEL sensors.
From a process point of view, the compatibility study of the pyroelectric frontplane with the TFT backplane has been achieved leading to significant process development and adaptation in the both TFT and pyroelectric components process flows.
PYCSEL 'one-finger' sensors design represented also extensive works during this period and resulted in providing all the necessary mask set tapes for the manufacturing of the first fingerprints sensors (first sensors are related to one finger capture; in the next period new design will be done related to a ‘4-finger’ capture imager).
The system integration, related to the development and manufacturing of the driving and readout electronics, has been also addressed during this first period and first monitoring tests of the sensors have been achieved leading to demonstrate the pyroelectric signal detection.
Main activities from M18 to M36:
- the driving and readout electronics of the 'one-finger' (256 x 256 pixels) sensor was finalized enabling the first acquisition demos and campaign.
- the specific design of the '4-finger' capture imager has been achieved taking into account the lessons learned from the 'one-finger' sensor realised during the first period. We created four different timing options for the readout limiting the number of I/Os and accelerating the acquisition,
- IGZO backplanes for the large area fingerprint sensors were next manufactured followed by the printing process of the large area pyroelectric sensor layers,
- the design and development of a new demoboard enabling the acquisition of 1500 x 1600 pixels has been necessary with specific throughput challenges solved by using USB3 interface and specific memory buffer,
- in the meantime, the quality of the 'one-finger' capture imager has been assessed and compared to a commercial optical imager, with good results. A specific filter for thermal images has been developed to fit the recognition algorithms. An acquisition campaign was performed and enabled to evaluate the recognition performance of the 'one-finger' sensor showing good results regarding the TRL level of the PYCSEL sensor. We also showed the low impact of a slight curvature of the fingerprint sensor on the image,
- Last, the PYCSEL consortium was awarded with the "best publicly funded project demonstrator" at the LOPEC OE-A competition 2019. A paper was also presented at the IEDM 2019 and an other one has been accepted at the SID 2020.
PROGRESS BEYOND THE STATE OF THE ART
Compared to other technologies, active thermal sensing offer a lot more advantages regarding integration ability, high quality image and ‘fake or difficult’ fingers detection. PYCSEL sensor technology lies on an innovative concept of flexible fingerprint sensor based on a smart design of pyroelectric active thermal pixels, using an IGZO TFT backplane on polyimide foil.
Today there is neither available nor in development active thermal pixels based on pyroelectric sensing. Based on thermal and electrical modelling of the pyroelectric device, a new smart design of the active thermal pixels will be proposed, enabling pyroelectric devices usable in static mode and on large area.
EXPECTED POTENTIAL IMPACT
Moving flexible and wearable electronics to a next level of applications addressing the growing worldwide biometrics market is the ambition of PYCSEL. The biometrics market is expected to be worth more than 20 billion € by 2020. The project is expecting impact not only in governmental market but also in high volume automotive (personalized HMIs), machine tool (user-restricted HMI), buildings (access control), consumer electronics and in enterprise market like banking applications, expected to be a huge lever in the future.
Compared to other technologies, active thermal sensing offer a lot more advantages regarding integration ability, high quality image and ‘fake or difficult’ fingers detection. PYCSEL sensor technology lies on an innovative concept of flexible fingerprint sensor based on a smart design of pyroelectric active thermal pixels, using an IGZO TFT backplane on polyimide foil.
Today there is neither available nor in development active thermal pixels based on pyroelectric sensing. Based on thermal and electrical modelling of the pyroelectric device, a new smart design of the active thermal pixels will be proposed, enabling pyroelectric devices usable in static mode and on large area.
EXPECTED POTENTIAL IMPACT
Moving flexible and wearable electronics to a next level of applications addressing the growing worldwide biometrics market is the ambition of PYCSEL. The biometrics market is expected to be worth more than 20 billion € by 2020. The project is expecting impact not only in governmental market but also in high volume automotive (personalized HMIs), machine tool (user-restricted HMI), buildings (access control), consumer electronics and in enterprise market like banking applications, expected to be a huge lever in the future.