The objectives of the Integrated Project e-CUBES are to advance the micro-system technologies to allow for the cost effective realisation of highly miniaturised, truly autonomous systems for ambient intelligence.

e-CUBES will reduce the costs, increase the performance and improve reconfigurability, scalability, adaptability and self-adjusting capabilities of complex systems in important economical application areas. In addition, the level of integration and miniaturisation will be increased allowing for improved interfacing with the surrounding and with networked services and systems. In particular, e-CUBES pursues the objectives of developing key micro-system technologies for the 3D integration of various heterogeneous layers, according to some key identified driving applications. e-CUBES will further demonstrate this through the realization of industrial prototypes with new functionalities and improved performance.

With its objectives e-CUBES clearly contributes to the objectives of IST in FP6: It ensures European leadership in the generic and applied technologies at the heart of the knowledge economy. e-CUBES increases innovation and competitiveness in European businesses and industry and to contributes to greater personal well-being of all European citizens.
e-CUBES serves to fulfil Focus 2 of 2.4.2 Technologies and devices for micro/nano-scale integration: Technology for very high density hybrid integration (towards e-grains, e-dust).
e-CUBES research activities address a family of integration and interfacing technologies aiming at very high densities, address unifying heterogeneous technologies including 3-dimensional vertical integration and very thin technologies. Furthermore in agreement with the IST objectives e-CUBES envisages integration of wireless communication interfaces, antennae, power provision and new functionalities into a very small volume/area.


The basic concept of "electronic cubes: e-CUBES", within the frame of global ambient intelligence, is shown in Figure 1. e-CUBES devices measure parameters in their environment (through a sensor function) and communicate this information, with rather low data rates, within an ad-hoc wireless network of e-CUBES and towards central nodes that connect the e-CUBES network to other network services and systems and to the user.

Figure 1: The concept of e-CUBES, within the frame of global ambient intelligence.

The considerations undertaken on the various foreseeable applications led the consortium to define a preliminary architecture described in Figure 2:

Figure 2: Prelimary architecture of e-CUBES

The reason for existence of these different functions are as follows:

  The sensors are the basic components of the e-CUBES. The preferred technology is MEMS, with a mandatory characteristics of ultra-low power, for instance using capacitive sensing.
 

The low power analogue interface is in charge of interfacing the analogue physical world seen by the sensors to the digital world of the microcontroller. The main constraints on this function is the minimisation of the internal components to reduce the power consumption, while being compatible with:
- noise levels adapted to the resolution of the sensors
- the necessary bandwidth
- the dynamic range
The technological challenge here, for an optimal version, is to identify a process compatible not only with all these constraints, but also with the digital and analogue components and basic sub-functions.

 
  The microcontroller is in charge of interfacing with pretty all the functions of the e-CUBES, particularly for sequencing all the operations, while controlling the power dissipation to reduce it at the lowest possible level. Its functionality is extended in some applications to data compression operations, self-organising procedure with other e-CUBES for data transmissions and relay, calibration operations, and other specific operations (such as localisation procedures of the other e-CUBES).
 
  The wireless interface is in charge of allowing the best link budget as possible for the communications, using a combination of techniques, from specific antenna schemes to dynamic impedance matching. This interface is bi-directional: uploading programs and commands from the base station, and getting data back from the sensors. The main challenge here is the optimal compromise between the highest efficiency of the transmissions, the lowest power dissipation during transmissions, and the reduced dimensions acceptable in a given domain.
 
  The events storage memory is a generic memory with specific properties of very low power consumed for a given amount of memory, data/state remanence, possibly dynamic programmability (re-programming), and dynamic sharing between data and programs. In the simplest case, it can be a standard flash memory. In the more complex case, it can be a nanocells memory array (millipede type for instance)
 
  The smart clock is a specific function useful for different cases. It can be used as:
- an "awakener" of the microcontroller in the case of ultra low power specific applications such as transient event monitoring where, most of the time, there is nothing (no event) to be noticed, but there are specific short time events appearing from time to time, for which a high resolution is necessary to identify the problems to be monitored.
- or a self rate-switchable clock (for power saving objectives).
- or a smart clock for localisation purposes (different possible and more or less complex procedures).
 
  The micropower source is the basic local energy source of any basic e-CUBES. In the simplest case, it can be a micro-battery or a supercapacitor. It can also be a micro-fuel cell in more elaborate versions. Its main characteristics is that its lifetime is necessarily limited in time, contrarily to the energy scavenger described hereunder.
 
  The energy scavenger is a function designed to get energy from the environment:
- whether from the real natural physical flows of energy (optical flux, EM signals,
thermal flux, ...).
- or whether from artificial sources, such as in the situation of remote powering, where a source of power is directed towards the e-CUBES from an external location.
Its intrinsic lifetime is not limited as a battery, at least up to the reliability level of the technology chosen to implement it. The complexity of this function can be high, depending on the possibility of having multiple sources of energy, as specific natural sources can vary in time, depending on the exact type of application (movements or not, day/night, ...).
 
  The power manager is a smart circuit able to handle the power management issues linked to the multiplicity of energy sources, including the local ones, by analysing temporally the available energy densities, deciding of a procedure for maximising the collected energy for the e-CUBES, by switching in time between the different energy sources, while storing this energy in a specific storage component (micro-battery as a buffer or capacitor, or any other mechanism including micro-mechanical energy storage devices).

The main common characteristics of all these functions is the very low power level for operating reliably (without interruption of the Service provided). It is also understood that, in the present state of the technology, the underlying technologies are heterogeneous, and that consequently, one key challenge is the heterogeneous integration of these functions, each realised optimally with its most suited technology.

In order to achieve a cost effective solution for the highly miniaturised e-CUBES system, the Integrated Project proposes the use of 3D interconnect technologies ("Cubic" interconnects - hence the name e CUBES), the use of modularity (reuse) and the use of wafer level fabrication technologies (in order to reach the required economics of scale). The e-CUBE is a 3D stack of functional sub-modules, each of which is, in itself, composed of a 3D stack of different (heterogeneous) functional layers (e.g. e-CUBE application layers). This is illustrated in (Figure 3).

Figure 3: Basic structure of an e-CUBES system: 3D stacking of functional sub-system layers

The minimal physical size of an e-CUBE is determined by three main factors: the size of the constituent integrated circuits, the minimum size of the antenna required for power-efficient communication, and the power storage and harvesting subsystem. The scaling roadmap of integrated circuit technologies enables the realisation of complex circuits in rather small sizes. The main size limitations are therefore the antenna size and the power system. Given the projected improvements in integrated circuit technology, with respect to die size, power consumption and frequency capabilities, the target size for e-CUBES in this project is to be smaller than 1cm3 (see Figure 4), targeting e-CUBES with dimensions 1mm*1mm*0.5mm in the future .

Figure 4: Roadmap for reduction in size of the e-CUBES