CitationDownload as .RIS
Emerald Group Publishing Limited
Copyright © 2009, Emerald Group Publishing Limited
Introducing AI into MEMS can lead us to brain-computer interfaces and super-human intelligence
Article Type: Viewpoint From: Assembly Automation, Volume 29, Issue 4
The author David Sanders is a Reader in Systems and Knowledge Engineering based in the Faculty of Technology, University of Portsmouth, Portsmouth, UK
Last year, I spoke about the progress being made in machine intelligence (Sanders, 2008c) and with sensors and networks of sensors (Sanders, 2008b). Earlier this year (in this journal) I spoke about ambient-intelligence, rapid-prototyping and the role of humans in the factories of the future (Sanders, 2009a). I addressed new applications and technologies such as merging machines with human beings, micro-electromechanics, electro-mechanical systems that can be personalized, smarter than human intelligence and swarms of smart sensors. Although the research to get us to all that will include human-machine interfaces, sensors, artificial intelligence (AI) and ambient intelligence (AmI), there is one technology above all others that has the potential to get us there fast […] and that is the creation and development of intelligent micro-electromechanical machines (MEMS).
The potential of very small machines was appreciated long before the technology existed to actually make them. For example, the Nobel Laureate (Feynman) considered the ability to manipulate matter on an atomic scale (Feynman, 1960) and he concluded his 1959 talk with two challenges: first for anyone to build a tiny motor and second for anyone to write the information from a book page onto a surface 1/25,000 smaller. He offered prizes of $1,000 for each (Feynman, 1960). The prize for the motor was won quickly using conventional tools but it was much later in 1985 when Tom Newman successfully reduced the first paragraph of “A Tale of Two Cities” and collected the second $1,000 prize (Gribbin, 1997).
MEMS only became practical once they could be fabricated using modified semiconductor fabrication technologies such as moulding and plating, wet and dry etching and electro-discharge machining. MEMS are made up of components between 1 and 100 μm in size and MEMS devices are smaller than a millimetre long. They usually consist of a central microprocessor and other components that can interact with the outside world such as micro-sensors (Waldner, 2008). At these sizes, the customary classical physics (that we are all used to) do not always hold true. Because of the much larger surface areas compared to volumes, surface effects such as electrostatics can dominate effects such as inertia or thermal mass. For example, gravity becomes less important and instead van der Waals attraction and surface tension can become more important.
MEMS devices will use less energy, space and time, and we will come to expect more and more output for less cost. Sensors and networks of sensors are already transforming manufacturing and assembly by scrutinizing our industrial environment and sometimes feeding into control systems to improve our processes. Individual sensors have tended to obtain data and then to transform that data into electrical signals to feed higher level systems (Sanders, 2008b). The development of such sensors has been driven by needs to reduce size and cost while increasing performance and MEMS could revolutionize the sensor markets by providing very small and reliable devices at minimal cost.
Until recently […] sensors tended to be simple, unintelligent, connected directly into control systems, and static (or at best moved from place to place by separate transportation systems) […] but all that is changing. Wireless networks are becoming increasingly common and some smaller sensors are becoming mobile so that networks of sensors can work in mobile teams (or swarms) (Sanders, 2008b). They can deploy and locate themselves around a factory or a machine to efficiently sample (and sometimes then control) the environment around them. Sensors are becoming “Smart Sensors” that can pre-process their own data to improve quality and reduce communications. These sensors become really smart when integral processing results in an adaptive sensing system that can react to external conditions and still provide useful measurements in harsh manufacturing and assembly conditions. Our future may be set to change through a combination of: smart mobile industrial sensors with enough energy to change themselves within their environment (for example, to move themselves around); effective wireless communication; automatic ranging; remote calibration; advances in microprocessors; new algorithms; and reduced costs in some key areas (Sanders, 2008b).
Increasing processing power within individual industrial sensors is improving the performance of sensor arrays and allowing for more accurate sensing of some phenomena that have traditionally required a large amount of off-line signal processing, such as image processing, sensor integration and gas sensor arrays (Sanders, 2008b). As the information from numbers of sensor arrays increases (and therefore becomes more complicated) then this leads to a need for systems to model and then convey the information in a simple way (and sometimes in real time) to human beings. These electronic sensor systems and their components are sometimes exposed to harsh environmental conditions and some new industrial sensors could be especially robust in harsh conditions. For example, some MEMS-sensors appear to be able to withstand very high humidity, pressure and temperature and these sensors-on-a-chip are potentially low cost (Sanders, 2008b).
Meanwhile, AI systems have been improving for a decade (Sanders, 2008c, 2009a, 1999) and AmI for assembly and manufacturing has been developing slowly (Sanders, 2009a; Sanders et al., 2008; Riva et al., 2005). These promise to bring improvements in flexibility, reconfigurability and reliability. Machine intelligence combines a wide variety of advanced technologies to give machines an ability to learn, adapt, make decisions and display new behaviours (Sanders, 2008c, 2009a). This is achieved using technologies such as neural networks (Sanders et al., 1996: Sanders, 2009b), expert systems (Hudson et al., 1997; Sanders and Hudson, 2000; Sanders et al., 2000; Tewkesbury and Sanders, 1999a), self-organizing maps (Sanders, 2008b; Burn and Home, 2008), fuzzy logic (Sanders, 2009a; Zoumponos and Aspragathos, 2008) and genetic algorithms (Sanders, 2008c; Manikas et al., 2007) and that machine intelligence technology has been developed through its application to many areas, such as: assembly (Sanders, 2009a, c; Schraft and Ledermann, 2003; Guru et al., 2004), building-modelling (Sanders, 2009a; Gegov, 2004; Wong et al., 2008), computer vision (Sanders, 2009c, 1993; Bertozzi et al., 2008; Chester et al., 2007, 2006; Sanders et al., 1992), environmental engineering (Sanders, 2008b; Hinks et al., 1996, 1995; Sanders et al., 2001, 1994; Hudson et al., 1996), human–computer interaction (Sanders et al., 2005; Sanders, 2009b; Sanders and Baldwin, 2001; Stott and Sanders, 2000b; Zhao et al., 2008), internet use (Bergasa-Suso et al., 2005; Kress, 2008), powered-wheelchair assistance (Sanders and Stott, 1999; Stott et al., 1997; Goodwin et al., 1997; Pei et al., 2007), maintenance and inspection (Nadakatti et al., 2008; Assembly Automation, 2008), medical systems (Stott and Sanders, 2000b; Sanders and Stott, 1999; Ohbayash, 2008), robotic manipulation (Urwin-Wright et al., 2003; Tewkesbury and Sanders, 1999a; Bullinaria and Li, 2007; Sreekumar et al., 2007), robotic programming (Tewkesbury and Sanders, 1999a, b, 1994, 2001; Urwin-Wright et al., 2002; Sanders and Rasol, 2001; Bogue, 2008) and sensing (Sanders, 2008a, b, 1999, 1993, 2007; Stott and Sanders, 2000a).
Our machines are exceeding human performance in more and more tasks (from guiding objects to assembling other machines) and some developments in machine intelligence are already being introduced into new manufacturing methods such as rapid-manufacture (Sanders, 2009a) and the manufacture of composites (Zhang and Richardson, 2007; Wang et al., 2006; Ferreira et al., 2007). If they can be effectively introduced into MEMS devices and into the manufacture of MEMS devices then machines can be made to merge with us more intimately and we should be able to combine our brain power with computer capacity to create a powerful AI. It is difficult to see the boundaries to what may be possible then and some scientists are predicting a period when the pace of technological change will be so fast and far-reaching that our lives will be irreversibly altered (Sanders, 2008c). At that point we may need a different type of engineer (Sanders and Harrison, 1992; Harrison and Sanders, 1992).
There are some interference problems that might become critical for wireless communications between MEMS and they can also be limited by antenna size, power and bandwidth, and that is all being explored by some radio engineers. MEMS will need the ability to cope with technology or communication failures and large-scale deployments and large amounts of data will need new computer science algorithms. Computer scientists are investigating some of that. The important and difficult future problems for MEMS may include constraints on resources such as energy, memory, computational speed and bandwidth. These limitations really push research towards distributed energy-efficiency. It is this potential need for smaller and more energy efficient sensors that can operate autonomously in harsh industrial conditions that will drive research towards more robust and fault tolerant MEMS that can automatically compensate for variables such as temperature.
For the immediate and medium term future, useful advances will come from research into: human-machine interfaces, sensors, AI, AmI, modelling and improving MEMS manufacturing and design techniques. In the longer term, understanding the properties of MEMS materials and then creating more capable and intelligent MEMS machines will lead to direct brain-computer interfaces that will allow us to communicate our ideas directly to machines (and to other human members of virtual teams) and that may change our world beyond recognition.
Assembly Automation (2008), “An intelligent inspection solution for smarter seating”, Assembly Automation, Vol. 28 No. 2, p. 173
Bergasa-Suso, J., Sanders, D.A. and Tewkesbury, G.E. (2005), “Intelligent browser-based systems to assist internet users”, IEEE Transactions, Vol. 48 No. 4, pp. 580–5
Bertozzi, M., Bombini, L., Broggi, A., Cerri, P., Grisleri, P., Medici, P. and Zani, P. (2008), “GOLD: A framework for developing intelligent-vehicle vision applications”, IEEE Intelligent Systems, Vol. 23 No. 1, pp. 69–71
Bogue, R. (2008), “Swarm intelligence and robotics”, Industrial Robot: An International Journal, Vol. 35 No. 6, pp. 488–95
Bullinaria, J.A. and Li, X.L. (2007), “An introduction to computational intelligence techniques for robot control”, Industrial Robot: An International Journal, Vol. 34 No. 4, pp. 295–302
Burn, K. and Home, G. (2008), “Environment classification using Kohonen self-organizing maps”, Expert Systems, Vol. 25 No. 2, pp. 98–114
Chester, S., Tewkesbury, G.E., Sanders, D.A. and Salter, P. (2006), “New electronic multi-media assessment system”, WEBIST 2006: Proceedings of the Second International Conference on Web Information Systems and Technologies, Setúbal, pp. 320–4
Chester, S., Tewkesbury, G.E., Sanders, D.A. and Salter, P. (2007), “New electronic multi-media assessment system”, Web Information Systems and Technologies, Vol. 1, pp. 414–20
Ferreira, J.M., Pires, J.T.B., Costa, J.D., Errajhi, O.A. and Richardson, M. (2007), “Fatigue damage and environment interaction of polyester aluminized glass fiber composites”, Composite Structures, Vol. 78 No. 3, pp. 397–401
Feynman, R.P. (1960), “There's plenty of room at the bottom: an invitation to enter a new field of physics”, Caltech's Engineering and Science, Vol. XXIII No. 5, pp. 22–36, available at: www.zyvex.com/nanotech/feynman.html
Gegov, A. (2004), “Application of computational intelligence methods for intelligent modelling of buildings: applications and science in soft computing”, in Loffi, A. and Garobaldi, J. (Eds), Advances in Soft Computing, Springer, Berlin, pp. 263–70
Goodwin, M.J., Sanders, D.A., Poland, G.A. and Stott, I.J. (1997), “Navigational assistance for disabled wheelchair-users”, Journal of Systems Architecture, Vol. 43 Nos 1/5, pp. 73–9
Gribbin, J. (1997), Richard Feynman: A Life in Science, Dutton, New York, NY, p. 170
Guru, S.M., Fernando, S., Halgamuge, S. and Chan, K. (2004), “Intelligent fastening with A-BOLT technology and sensor networks”, Assembly Automation, Vol. 24 No. 4, pp. 386–93
Harrison, D.J. and Sanders, D.A. (1992), “Imagineering – promoting creativity in engineering education”, Engineering Education, Vol. 2, pp. 45–9
Hinks, J.W., Cawte, H., Sanders, D.A., Hudson, A. and Dockree, C. (1996), “Prediction of flow rates and stability in large scale airlift reactors”, Water Science and Technology, Vol. 34 Nos 5/6, pp. 51–7
Hinks, J.W., Cawte, H., Sanders, D.A., Fenske, O., Poland, G.A. and Tewkesbury, G.E. (1995), “Model for the prediction of liquid volumetric flow rates in large scale airlift reactors”, 3rd International Conference on Water and Waste Treatment, BHR Group Conference Series, No. 17, BHR Group, Cranefield, pp. 125–33
Hudson, A.D., Sanders, D.A., Golding, H., Tewkesbury, G.E. and Cawte, H. (1997), “Aspects of an expert design system for the wastewater treatment industry”, Journal of Systems Architecture, Vol. 43 Nos 1/5, pp. 59–65
Hudson, A.D., Sanders, D.A., Tewkesbury, G.E., Cawte, H. and Hinks, J. (1996), “Simulation of a high recirculation airlift reactor for steady-state operation”, Water Science and Technology, Vol. 34 Nos 5/6, pp. 59–66
Kress, M. (2008), “Intelligent internet knowledge networks: processing of concepts and wisdom”, Information Processing & Management, Vol. 44 No. 2, pp. 983–4
Manikas, T.W., Ashenayi, K. and Wainwright, R.L. (2007), “Genetic algorithms for autonomous robot navigation”, IEEE Instrumentation & Measurement, Vol. 10 No. 6, pp. 26–31
Nadakatti, M., Ramachandra, A. and Kumar, A.N.S. (2008), “Artificial intelligence-based condition monitoring for plant maintenance”, Assembly Automation, Vol. 28 No. 2, pp. 143–50
Ohbayash, K. (2008), “Advanced cancer research with PXI and LabVIEW”, Instrumentation Newsletter, Q3 p. 29
Pei, J., Huosheng, H., Tao, L. and Kui, Y. (2007), “Head gesture recognition for hands-free control of an intelligent wheelchair”, Industrial Robot: An International Journal, Vol. 34 No. 1, pp. 60–8
Riva, G., Vatalaro, F., Davide, F. and Alcañiz, M. (2005), Ambient Intelligence in Practice: Future Perspectives and Applications, IOS Press, Amsterdam, pp. 237–64
Sanders, D.A. (1993), “System specification 2”, Microprocessing and Microprogramming, Vol. 38 Nos 1/5, pp. 833–4
Sanders, D.A. (1999), “Perception in robotics”, Industrial Robot: An International Journal, Vol. 26 No. 2, pp. 90–2
Sanders, D.A. (2007), “Viewpoint – force sensing”, Industrial Robot: An International Journal, Vol. 34 No. 4, p. 268
Sanders, D.A. (2008a), “Controlling the direction of ‘walkie’ type forklifts and pallet jacks on sloping ground”, Assembly Automation, Vol. 28 No. 4, pp. 317–24
Sanders, D.A. (2008b), “Environmental sensors and networks of sensors”, Sensor Review, Vol. 28 No. 4, pp. 273–4
Sanders, D.A. (2008c), “Progress in machine intelligence”, Industrial Robot: An International Journal, Vol. 35 No. 6, pp. 485–7
Sanders, D.A. (2009a), “Ambient intelligence and energy efficiency in rapid prototyping and manufacturing”, Assembly Automation, Vol. 29 No. 3 (in press)
Sanders, D.A. (2009b), “A pointer device for TFT display screens that determines position by detecting colours on the display using a colour sensor and an artificial neural network”, Displays, Vol. 30 No. 2, Displays Paper No. DISPLA-D-08-00006, pp. 84–96
Sanders, D.A. (2009c), “Recognizing shipbuilding parts using artificial neural networks and Fourier descriptors”, Proc. IMechE, Part B: Journal of Engineering Manufacture, Vol. 223, B3, pp. 337–42
Sanders, D.A. and Baldwin, A. (2001), “X-by-wire technology”, Total Vehicle Technology: Challenging Current Thinking, Department of Mechanical and Manufacturing Engineering, University of Portsmouth, Portsmouth, pp. 3–12
Sanders, D.A. and Harrison, D.J. (1992), “The requirement for integrated and systems engineers in a complex and flexible future”, Engineering Education, Vol. 3, pp. 343–8
Sanders, D.A. and Hudson, A.D. (2000), “A specific blackboard expert system to simulate and automate the design of high recirculation airlift reactors”, Mathematics and Computers in Simulation, Vol. 53 Nos 1/2, pp. 41–65
Sanders, D.A. and Rasol, Z. (2001), “An automatic system for simple spot welding tasks”, Total Vehicle Technology: Challenging Current Thinking, Department of Mechanical and Manufacturing Engineering, University of Portsmouth, Portsmouth, pp. 263–72
Sanders, D.A. and Stott, I.J. (1999), “A new prototype intelligent mobility system to assist powered wheelchair users”, Industrial Robot: An International Journal, Vol. 26 No. 6, pp. 466–75
Sanders, D.A., Cawte, H. and Hudson, A.D. (2001), “Modelling of the fluid dynamic processes in a high-recirculation airlift reactor”, International Journal of Energy Research, Vol. 25 No. 6, pp. 487–500
Sanders, D.A., Harris, P. and Mazharsolook, E. (1992), “Image modelling in real time using spheres and simple polyhedra”, 4th Int. Conf. on Image Processing and its Applications, Manchester, Vol. 354, pp. 433–6
Sanders, D.A., Haynes, B.P., Tewkesbury, G.E. and Stott, I.J. (1996), “The addition of neural networks to the inner feedback path in order to improve on the use of pre-trained feed forward estimators”, Mathematics and Computers in Simulation, Vol. 41 Nos 5/6, pp. 461–72
Sanders, D.A., Hudson, A.D., Tewkesbury, G.E. and Mason, S. (2000), “Automating the design of high-recirculation airlift reactors using a blackboard framework”, Expert Systems with Applications, Vol. 18 No. 3, pp. 231–45
Sanders, D.A., Liu, H., Harrison, D.J. and Gegov, A. (2008), “Energy efficiency based on ambient intelligence”, Journal of Computing in Systems and Engineering, Vol. 9 No. 1, pp. 114–20
Sanders, D.A., Urwin-Wright, S.D., Tewkesbury, G.E. and Gremont, B. (2005), “Pointer device for thin-film transistor and cathode ray tube computer screens”, Electronics Letters, Vol. 41 No. 16, pp. 894–6
Sanders, D.A., Hudson, A.D., Cawte, H., Fenske, O., Poland, G.A. and Tewkesbury, G.E. (1994), “Computer modelling of single sludge systems for the computer-aided design and control of activated sludge processes”, Microprocessors and Microprogramming, Vol. 40 Nos 10/12, pp. 867–70
Schraft, R.D. and Ledermann, T. (2003), “Intelligent picking of chaotically stored objects”, Assembly Automation, Vol. 23 No. 1, pp. 38–42
Sreekumar, M., Nagarajan, T., Singaperumal, M., Zoppi, M. and Molfino, R. (2007), “Critical review of current trends in shape memory alloy actuators for intelligent robots”, Industrial Robot: An International Journal, Vol. 34 No. 4, pp. 285–94
Stott, I.J. and Sanders, D.A. (2000a), “New powered wheelchair systems for the rehabilitation of some severely disabled users”, International Journal of Rehabilitation Research, Vol. 23 No. 3, pp. 149–53
Stott, I.J. and Sanders, D.A. (2000b), “The use of virtual reality to train powered wheelchair users and test new wheelchair systems”, International Journal of Rehabilitation Research, Vol. 23 No. 4, pp. 321–6
Stott, I.J., Sanders, D.A. and Goodwin, M.J. (1997), “A software algorithm for the intelligent mixing of inputs to a tele-operated vehicle”, Journal of Systems Architecture, Vol. 43 Nos 1/5, pp. 67–72
Tewkesbury, G.E. and Sanders, D.A. (1994), “The automatic programming of production machinery for de-flashing plastic parts”, Advances in Manufacturing Technology, Vol. VIII, pp. 279–83
Tewkesbury, G.E. and Sanders, D.A. (1999a), “A new robot command library which includes simulation”, Industrial Robot: An International Journal, Vol. 26 No. 1, pp. 39–48
Tewkesbury, G.E. and Sanders, D.A. (1999b), “A new simulation based robot command library applied to three robots”, Journal of Robotic Systems, Vol. 16 No. 8, pp. 461–9
Tewkesbury, G.E. and Sanders, D.A. (2001), “The use of distributed intelligence within advanced production machinery for design applications”, Total Vehicle Technology: Challenging Current Thinking, Department of Mechanical and Manufacturing Engineering, University of Portsmouth, Portsmouth, pp. 255–62
Urwin-Wright, S., Sanders, D.A. and Chen, S. (2002), “Terrain prediction for an eight-legged robot”, Journal of Robotic Systems, Vol. 19 No. 2, pp. 91–8
Urwin-Wright, S., Sanders, D.A. and Chen, S. (2003), “Predicting terrain contours using a feed-forward neural network”, Engineering Applications of Artificial Intelligence, Vol. 16 Nos 5/6, pp. 465–72
Waldner, J.B. (2008), Nanocomputers and Swarm Intelligence, ISTE, London
Wang, W., Song, M., Zhang, Z.Y. and Richardson, M. (2006), “Synthesis and characterization of high nickel-containing mesoporous silica via a modified direct synthesis method”, Journal of Non-Crystalline Solids, Vol. 352 Nos 21/22, pp. 2180–6
Wong, J., Li, H. and Lai, J. (2008), “Evaluating the system intelligence of the intelligent building systems”, Automation in Construction, Vol. 17 No. 3, pp. 303–21
Zhang, Z.Y. and Richardson, M.O.W. (2007), “Low velocity impact induced damage evaluation and its effect on the residual flexural properties of pultruded GRP composites”, Composite Structures, Vol. 81 No. 2, pp. 195–201
Zhao, M., Nowatzyk, A.G., Lu, T. and Farkas, D.L. (2008), “Intelligent non-contact surgeon-computer interface using hand gesture recognition. Three dimensional image capture and applications”, Proceedings of the SPIE-IS&T Electronic Imaging, San Jose, CA, Vol. 6805, p. U8050
Zoumponos, G.T. and Aspragathos, N.A. (2008), “Fuzzy logic path planning for the robotic placement of fabrics on a work table”, Robotics & Computer-Integrated Manufacturing, Vol. 24 No. 2, pp. 174–86
Sanders, D.A. (1995), “Real time geometric modelling using models in an actuator space and Cartesian space”, Journal of Robotic Systems, Vol. 12 No. 1, pp. 19–28
Sanders, D.A. (1995), “The modification of pre-planned manipulator paths to improve the gross motions associated with the pick and place task”, ROBOTICA, Vol. 13, pp. 77–85