Sensor-guided micro assembly of active micro systems by using a hot melt based joining technology

This article presents (the most) recent results of the subprojects B4 and B8 of the Collaborative Research Center 516—Design and Manufacturing of Active Micro Systems—which are concerned with the assembly of active micro systems. While subproject B4 investigates sensor guided assembly processes, subproject B8 develops suitable assembly techniques on the basis of non-viscous adhesive systems (hot melts). Process development focuses on the suitability for automation, process times and the applicability of batch processes. The article discusses certain hot melt application techniques that are suitable for batch production, a sensor-guided assembly system as well as different approaches for heat conduction in an automated assembly process for hot melt coated micro components.     

Assembly of hybrid integrated micro-optical modules using passive alignment with LIGA mounting elements and adhesive bonding techniques

 Assembling hybrid integrated micro-optical modules for mono-mode applications requires alignment tolerances better than ±1 μm which, so far, were achieved by complex active alignment of the micro-optical components. This article describes a passive assembly concept employing high-precision LIGA structures to simplify assembly and reduce manufacturing costs. The advantages of this passive assembly concept are demonstrated in the micro-optical assembly of a heterodyne receiver. The micro-optical components (ball microlenses, prisms, glass fibers, photodiodes) were aligned passively by means of alignment structures made of PMMA, and were subsequently fixed by UV-curing adhesive. Photodiodes were additionally contacted with electrically conducting adhesive cured at 70 °C. The mono-mode glass fibers were mounted in fiber mounts into the fiber grooves of which the glass fibers were inserted and immobilized with UV-curing adhesive. Measurements of the optical quality of the heterodyne receiver indicated an accuracy of assembly better than 1 μm. Received: 3 January 2000/Accepted: 9 February 2000     

Hybrid universal handling systems for micro component assembly

Highly sophisticated machining techniques make it possible to manufacture ultra precise micro components as small as 100 μm3. These micro components are normally mounted onto larger macroscopic components before being incorporated into, e.g., medical, telecommunication or sensor technology products. In view of the highly inflexible systems currently available, the work carried out by the Fraunhofer IPT with funding from the SFB 440 focuses on the development of a hybrid, universally applicable fine positioning system. The system consists of an assembly head which, with the aid of an integrated six-axis fine positioning unit and an integrated sensor unit for referencing the assembly position, gives a conventional positioning system such as an articulated robot arm the ability to perform highly precise micro assembly processes. This provides the context for the development work carried out by the Fraunhofer IPT into assembly head design concepts. The first step involved investigating the positioning behavior of a conventional articulated robot arm, and then analyzing the demands placed on the assembly head. This is followed by the development and optimization of the miniaturized guidance, actuator and sensor components that are needed for the construction of a robust, highly precise six-axis fine adjustment system. The result of this work includes a set of linear guides, the size of a matchbox, with air bearings. The prototype guides set up at the Fraunhofer IPT have a non-contact interface, which makes it possible to convey compressed air without friction or force.     

Fuzzy logic and position sensing for precision assembly

The problem of part mating and assembly with close tolerances has been addressed in the past by active or passive compliance and by force/position control. With ambient sensor and transmission noise and with imprecise measurements it is often difficult to attain high precision in manipulator positioning. In this article, a position-sensing wrist using simple strain gauges is described. A fuzzy logic algorithm to ascertain the positional error during an unsuccessful assembly attempt is explained. The results obtained by using this wrist and the fuzzy logic algorithm to reposition a manipulator for precision assembly are presented and discussed. © 1995 John Wiley & Sons, Inc.     

Fine motion strategy for skill-based manipulation

Generally, a manipulator task can be divided into several motion primitives called “skills.” Skill-based motion planning is an effective way to execute a complicated task. When planning an assembly process, a technique of fine motion planning such as the backprojection method in configuration space is often used. This paper describes fine motion planning using a skill library, which consists of a pattern of trajectories of skill motions in configuration space. This method gives the initial position and orientation from which the object can reach the goal in skill-based manipulation.     

ANN modeling of a smart MEMS-based capacitive humidity sensor

This paper presents a design of a smart humidity sensor. First we begin by the modeling of a Capacitive MEMS-based humidity sensor. Using neuronal networks and Matlab environment to accurately express the non-linearity, the hysteresis effect and the cross sensitivity of the output humidity sensor used. We have done the training to create an analytical model CHS “Capacitive Humidity Sensor”. Because our sensor is a capacitive type, the obtained model on PSPICE reflects the humidity variation by a capacity variation, which is a passive magnitude; it requires a conversion to an active magnitude, why we realize a conversion capacity/voltage using a switched capacitor circuit SCC. In a second step a linearization, by Matlab program, is applied to CHS response whose goal is to create a database for an element of correction “CORRECTOR”. After that we use the bias matrix and the weights matrix obtained by training to establish the CHS model and the CORRECTOR model on PSPICE simulator, where the output of the first is identical to the output of the CHS and the last correct its nonlinear response, and eliminate its hysteresis effect and cross sensitivity. The three blocks; CHS model, CORRECTOR model and the capacity/voltage converter, represent the smart sensor.     

Potential of 5,10,15,20-Tetrakis(3′,5′-di-tertbutylphenyl)porphyrinatocopper(II) for a multifunctional sensor

An organic compound 5,10,15,20-Tetrakis(3′,5′-di-tertbutylphenyl)porphyrinatocopper(II) (TDTPPCu) is synthesized and studied as an active material for multifunctional capacitive sensor. The capacitance of the device as a function of illumination, humidity and temperature has been investigated. It is observed that the capacitance increases by 4.7 times from the dark condition under an illumination of 3850 lx. The capacitance is also changed 9.5 times with the increase in relative humidity (RH) from 30% to 95%. No change in capacitance appeared below critical temperature 120 °C. Based on the experimental results for the multifunctional sensor a mathematical model has been developed. The model is mainly based on the assumption that the capacitive response of the sensor is associated with dielectric polarization. The sensors are simulated using this model. The simulated results match well with experimental results.     

Injection molding for microstructures controlling mold-core extrusion and cavity heat-flux

 In this work we constructed an injection press molding system with a mold-core extrusion mechanism and a small sensor assembly for effectively duplicating microstructures to the mold products. The mold-core extrusion mechanism is driven by a piezo element to apply force on important area with microstructures. For example, after injection it increases the cavity pressure from 20 to 34 MPa. Small sensors consist of the pressure, displacement, and heat flux sensor assemblies, arranged around the small cavity. The signals showed us the physical phenomena inside the mold and may be further used as control signal. In order to evaluate this injection press molding system, we formed micro triangular grooves of pitch 1 μm and angle 140°. The mold-core extrusion gave better diffraction intensity by several percents. Received: 16 May 2001/Accepted: 24 July 2001 The authors would like to thank FANUC Ltd. for cutting the triangular grooves on the mold-core. This paper was presented at the Conference of Micro System Technologies 2001 in March 2001.     

Micromachined gas sensors for environmental pollutants

An Au doped tin oxide thin film was deposited as base material for carbon monoxide detection over a micromachined substrate. The performances of a recent technique to heat the device, named fast pulsed temperature supply, are presented. This technique exploits the property that, due to the very low thermal mass of the membrane, the term required to reach steady state conditions is very short (about 40 ms). The sensor heater is periodically supplied for very short terms, hundred of milliseconds, and kept off for long ones, seconds or more. Besides a strong reduction of power consumption compared with isothermal characterization, an increase of sensitivity is observed. Different shapes of the heating wave were examined and results are summarized and compared. Received: 1 June 1999 / Accepted: 2 June 1999     

Microsensor packaging

 Packaging is crucial for the success of microsensors and microsystems, and typically a major, if not dominating cost component. Benefiting from packaging methods for integrated circuits (IC), microsystem packaging currently strongly relies on customized solutions. The paper summarizes challenges in microsensor and microsystem packaging and briefly discusses current packaging trends. In the second part of the paper, a packaged low-cost CMOS thermal imager is presented. Its packaging is based on the direct attachment of a silicon infrared filter onto the CMOS sensor die. The final microsystem is further packaged in a plastic ball grid array (BGA) enabling subsequent assembly by standard surface mount technology. Received: 20 July 2001 / Accepted: 23 July 2001     

A non-contact laser speckle sensor for the measurement of robotic tool speed

A non-contact speckle correlation sensor for the measurement of robotic tool speed is described that is capable of measuring the in-plane relative velocities between a robot end-effector and the workplace or other surface. The sensor performance has been assessed in the laboratory with sensor accuracies of ±0.01 mm/s over a ±70 mm/s velocity range. The effect of misalignment of the sensor on the robot was assessed for variation in both working distance and angular alignment with sensor accuracy maintained to within 0.025 mm/s (<0.04%) over a working distance variation of ±5 mm from the sensor design distance and ±0.4 mm/s (0.6%) for a misalignment of 5°. The sensor precision was found to be limited by the peak fitting accuracy used in the signal processing with peak errors of ±0.34 mm/s. Finally an example of the sensor’s application to robotic manufacturing is presented where the sensor was applied to tool speed measurement for path planning in the wire and arc additive manufacturing process using a KUKA KR150 L110/2 industrial robot.     

Ultra thin ICs and MEMS elements: techniques for wafer thinning,stress-free separation,assembly and interconnection

 Ultra thin chips with a thickness below 30 μm offer low system height, low topography and show enhanced mechanical flexibility. These properties enable diverse use possibilities and new applications. However, advanced wafer thinning, adapted assembly and interconnection methods are required for this technology. A new process scheme is proposed that allows manufacturing of ultra thin fully processed wafers. Secure handling is achieved by means of carrier substrates using reversible adhesive tapes for connection of support and device wafers. Well established backgrinding and etching techniques are used for wafer thinning. To avoid mechanical damage of thin ICs the “Dicing-by-Thinning” (DbyT) concept is introduced to process flow. Best results are obtained when preparing dry etched chip grooves at front side of device wafer and opening these trenches during backside thinning. The new process scheme was also applied to wafers with highly topographic surfaces. Results of 40 μm thin wafers with 15 μm high Nickel bumps are presented. Three different assembly methods are described, interconnection through the thin chip, face down assembly and isoplanar contacting. Received: 6 July 2001/Accepted: 26 February 2002 The authors would like to thank M. Küchler (IZM Chemnitz) for preparing and performing trench etching process and A. Ostmann (IZM Berlin) for performance of nickel bumping process. This paper was presented at the Conference of Micro System Technologies 2001 in March 2001.     

Fuzzy rule base optimization of a compliant wrist sensor for robotic assembly

The problem of part mating and assembly with close tolerances has been addressed in the past by active or passive compliance and by force/position control. With ambient sensor and transmission noise and with imprecise measurements it is often difficult to attain high precision in manipulator positioning. A compliant wrist position sensor with a fuzzy logic rule base was designed by the author to address this problem. In this article, after a brief review of this sensor, optimization of its fuzzy rule base is discussed. The optimization results are then presented. © 1996 John Wiley & Sons, Inc.     

Design of microfluidic channels separated by an ultra-thin free-standing dielectric membrane

A microfluidic (MF) surface plasmon polariton sensor featuring a gold Mach–Zehnder interferometer on an ultra-thin (20–35 nm) dielectric membrane is described. While the presence of the membrane is required to maintain a near mirror symmetry of the dielectric properties of the medium on either side of the interferometer, it is a source of unique challenges in the MF system design. The pressure required to drive the fluid flow in microchannels causes deflection whose value depends on the membrane’s residual stress in the low pressure range and on its modulus at the higher pressure range. Depending on the empirical membrane strength which would meet the required equipment reliability, narrow fluidic channels may require tight dimensional tolerances to maintain the pressure difference across the membrane below a critical value. With wider channels (≥100 μm) dimensional tolerances are relaxed even with relatively weak membranes.     

Microcoil speed and position sensor for automotive applications

 An inductive ABS speed and position sensor has been realized, based on an arrangement of batch-fabricated microcoils in a differential transformer configuration. The position of a metallic toothed wheel is measured with a precision of ±2 μm in a 0 to 3.2 kHz bandwidth and at a distance of 0 to 2 mm, with a transformer power consumption of less than 50 mW. Both the differential sensor approach and the ratiometric analysis allow for position detection that is insensitive to target speed, target distance and sensor temperature. The AC excitation allows the frequency to be adapted to the target material and to measure down to zero speed. Received: 1 November 1996/Accepted: 13 November 1996     

Fabrication of wireless sensors on flexible film using screen printing and via filling

Wireless sensors are fabricated on flexible plastic films by means of screen printing and via-hole filling. The wireless sensors are battery free with data and power transmission functions. The sensors, fabricated on polyethylene terephtalate films, are designed based on RFID technology. Using an additive patterning process known as screen printing, metallization on polymer films is created. Both sides of a polymer film are printed with metallic patterns and connected with micro vias filled with conductive paste. One side of the film consists of printed electrical traces for discrete components like resistors and transistors that would be mounted onto it; the other side consists of a printed inductive coil used for wireless data and power transmission. The micro vias, which have a diameter of 120 μm, are formed by mechanical punching and filled with conductive silver paste. The size of one sensor unit is approximately 2 cm × 1.5 cm; an array of 4 × 7 sensor units are printed over an area of 15 cm × 15 cm on a PET film. Details of manufacturing processes, component assembly and functionality test are presented in this paper.     

Similarity-based analysis for large networks of ultra-low resolution sensors

By analyzing the similarities between bit streams coming from a network of motion detectors, we can recover the network geometry and discover structure in the human behavior being observed. This means that a low-cost network of sensors can provide powerful contextual information to building systems: improving the efficiency of elevators, lighting, heating, and cooling; enhancing safety and security; and opening up new opportunities for human-centered information systems. This paper will show how signal similarity can be used to calibrate a sensor network to accuracies below the resolution of the individual sensors. This is done by analyzing the similarity structures in the unconstrained movement of people in the observed space. We will also present our efficient behavior-learning algorithm that yields 90% correct behavior-detection in data from a sensor network comprised of motion detectors by employing similarity-based clustering to automatically decompose complex activities into detectable sub-classes.     

Distributed agreement in tile self-assembly

Laboratory investigations have shown that a formal theory of fault-tolerance will be essential to harness nanoscale self-assembly as a medium of computation. Several researchers have voiced an intuition that self-assembly phenomena are related to the field of distributed computing. This paper formalizes some of that intuition. We construct tile assembly systems that are able to simulate the solution of the wait-free consensus problem in some distributed systems. (For potential future work, this may allow binding errors in tile assembly to be analyzed, and managed, with positive results in distributed computing, as a “blockage” in our tile assembly model is analogous to a crash failure in a distributed computing model.) We also define a strengthening of the “traditional” consensus problem, to make explicit an expectation about consensus algorithms that is often implicit in distributed computing literature. We show that solution of this strengthened consensus problem can be simulated by a two-dimensional tile assembly model only for two processes, whereas a three-dimensional tile assembly model can simulate its solution in a distributed system with any number of processes.     

MicroGP—An Evolutionary Assembly Program Generator

This paper describes μGP, an evolutionary approach for generating assembly programs tuned for a specific microprocessor. The approach is based on three clearly separated blocks: an evolutionary core, an instruction library and an external evaluator. The evolutionary core conducts adaptive population-based search. The instruction library is used to map individuals to valid assembly language programs. The external evaluator simulates the assembly program, providing the necessary feedback to the evolutionary core. μGP has some distinctive features that allow its use in specific contexts. This paper focuses on one such context: test program generation for design validation of microprocessors. Reported results show μGP being used to validate a complex 5-stage pipelined microprocessor. Its induced test programs outperform an exhaustive functional test and an instruction randomizer, showing that engineers are able to automatically obtain high-quality test programs.     

Direct-vision-based reinforcement learning in a real mobile robot

It was confirmed that a real mobile robot with a simple visual sensor could learn appropriate motions to reach a target object by direct-vision-based reinforcement learning (RL). In direct-vision-based RL, raw visual sensory signals are put directly into a layered neural network, and then the neural network is trained using back propagation, with the training signal being generated by reinforcement learning. Because of the time-delay in transmitting the visual sensory signals, the actor outputs are trained by the critic output at two time-steps ahead. It was shown that a robot with a simple monochrome visual sensor can learn to reach a target object from scratch without any advance knowledge of this task by direct-vision-based RL. This work was presented in part at the 7th International Symposium on Artificial Life and Robotics, Oita, Japan, January 16–18, 2002