The International Conference on Automation, Robotics and Artificial Intelligence Applied to Analytical Chemistry and Laboratory Medicine has become the premier meeting dedicated to expanding the knowledge base in the rapidly evolving field of laboratory automation. This year was the fifth consecutive year for the International Conference, and we experienced impressive growth in registrations and exhibitions.
The following pages review the conference, which was held at the Sheraton Hotel and Marina in San Diego, California this past January. Featured were 51 exhibitors, 14 sponsors, 54 oral presentations, 41 posters and over 600 registrants and exhibitor associates. The exponential growth in interest in this conference is fueled by the need to increase the diversity of tests available by laboratory analysis while, at the same time, reducing the labor necessary to produce quality results. Simultaneous improvements in quality, diversity and efficiency can only be achieved through robotic automation.
The success of the International Conference has also hinged on the fact that we bring together a wide variety of technologies that are immediately useful to the end user. Technology that will assure the financial success of laboratories in the future is of particular interest to conference participants. Automation is currently available for laboratory medicine, pharmaceutical high-throughput screening, environmental pollutant analysis, and analytical chemistry.
The five-day conference consisted of three presentation formats. On Monday, January 16 and Tuesday, January 17 short courses on selected emerging technology areas were taught so that novices could become familiar with the tools of automation. Among the offerings were courses on machine vision, electronic laboratory notebooks, use of the Internet, machine intelligence, laboratory information systems (LIMS), and the fundamentals of laboratory robotics and automation. The scientific program took place Wednesday through Friday, January 18–20, and consisted of oral presentations by authorities in a wide range of automation fields. From Tuesday through Friday, poster presentations were given in the exhibit hall area.
SCIENTIFIC PROGRAM: DAY ONE
Dr. Bart Kosko, 1995 Beckman Award Lecturer
The scientific program opened with a special session on artificial intelligence (or machine intelligence) presented by internationally recognized experts in their fields. Bart Kosko was featured as the Beckman Award Lecturer based on his international reputation in fuzzy logic. Dr. Kosko lectured about the application of fuzzy rules to laboratory problems. According to Kosko, fuzzy logic can be applied to solving laboratory problems in much the same manner as other machine intelligence techniques such as neural networks, genetic algorithms, and expert systems. However, fuzzy logic is most beneficial when the output is non-quantized black and white. The science of fuzzy logic concentrates on providing rules that contain the output within tolerable bounds.
Leon Klatt spoke about the use of artificial intelligence (AI) to assist in pattern recognition. He used specific examples of pattern recognition in environmental analysis. He explained that in pattern recognition one can use parametric as well as non-parametric techniques which involve learning and classification. Bobi Den Hartog discussed various linear and nonlinear pattern recognition techniques for interpreting gas chromatograms of PCBs. The presentation covered Multiple Linear Regression, Principle Component Regression, Peak Area Analysis and Neural Networks. Art White explained that almost one third of all telephone calls to companies that sell medical diagnostic instruments are related to the systematic trouble shooting of instrument problems. He is involved with the creation of a machine intelligence system to aid in the diagnosis and treatment of instrument failures. His group chose a decision tree based system since it was the most likely to solve instrument problems (the system they have created has correctly diagnosed 47 of the 50 test cases). David Brubaker presented a philosophical discussion about the difficulties of implementing fuzzy logic into industrial applications. As an industrial software engineer, he will use fuzzy logic if it solves a problem better than existing technology, if it is cost effective to implement, and if the client trusts the output data from a system that contains fuzzy logic code.
Randy Roberts explained the development of an automated qualitative system for the assessment of gas chromatograms. Similarly, Susan Hruska discussed the use of hybrid artificial intelligence technology in the automation of fault analysis in gas chromatography. Taiwei Lu's presentation dealt with the use of neural networks for real-time process control and instrument fault detection in analytical instruments. Alain Laugier from Technidata (Grenoble, France) described the use of neural networks as an intelligent component of their innovative Laboratory Information System (LIS). Alain demonstrated the classification of leukemia with only a 15% error rate.
OVERVIEW STATE-OF-THE-ART SESSION
Following the discussions of artificial intelligence, a series of investigators described their experience in automating the clinical laboratory. Al Kolb discussed strategies for automation of high-throughput screening. He suggested that selecting compounds based on structural diversity, testing multiple compounds per well and testing samples in one concentration are ways to reduce the number of compounds being tested or the number of microplate wells needed for testing. Jan van der Greef gave a talk on novel tools for the automated analytical laboratory, in which he described a dynamic analytical system for urine testing which uses rapid sub-nanoliter analysis coupled with an automated loading device. Georg Hoffmann (Trillium, Grafrath, Germany) represents a company that is establishing successful partnerships between technology, industry and the clinical laboratory. His lecture centered on explaining the relationships between Hospital Information Systems, Laboratory Information Systems and analytical instruments. Specialized data management systems will be required to form a bridge between instruments and automated devices and the established base of Laboratory Information Systems. His company plans to facilitate the path for industry to develop an open and scalable architecture to allow easy and immediate interconnectivity of information systems.
STANDARDIZATION FOCUS SESSION
At the end of the first day of lectures, the program focused on the critical issues surrounding standardization. Without standardization, many robotic systems will reach technological obsolescence in a short period of time. Standardization initiatives at the National Committee for Clinical Laboratory Standards were summarized by Charles Galanaugh. He described the recently sanctioned steering committee headed by Rodney Markin.
Peter Grandsard, a representative from the National Institutes of Standards and Technology, focused on the efforts of the Consortium on Automated Analytical Laboratory Systems. Grandsard presented in place of Gary Kramer who was unable to attend. The European clinical laboratory groups that are interested in standardization were summarized by Pierangelo Bonini.
FINAL DAY OF SCIENTIFIC SESSIONS: MICROFABRICATION TECHNOLOGIES
Microfabrication technology is creating mechanical devices that measure in the millimeter and submillimeter range. Micromotors will drive microdevices which will be able to perform many analytical tasks currently performed by analyzers many orders of magnitude larger. The microfabrication session is aimed at familiarizing registrants with existing technologies and forecasting the development of the micromachines that will make the micro laboratory systems of tomorrow.
Microfabrication of microwells for microreactions can now be performed routinely; however, filling the microwells with reagents requires special pipetting devices. David Wallace (Microfab Technologies Inc., Piano, TX) demonstrated the use of ink jet technology to create reproducible droplets that can be sprayed into an array of reaction chambers. He also demonstrated that the droplets that are formed by ink jet can be combined in transit so that reactions can be initiated before they arrive at the reaction well. Drops can be produced at 20,000/sec with a diameter of 20 μrn- 500 μm (4pL- 4 nL).
Nik Willmore discussed a method to produce many microwells in a non-wetting substrate. The microwells can be filled with the ink jet technology previously described or filled by dipping the substrate into a liquid. When removing the substrate small reproducible amounts of liquid remain in the wells. This technology could form reagent-filled wells which could become the basis of small analytical systems.
Ernest Garcia defined the field of microfabrication as an interdisciplinary field of science and technology concerned with the design and development of microelectromechanical systems. Microfabricated machines are small in size and can be mass produced at low cost, creating vast arrays of devices which can perform functions similar to integrated electrical circuits. For example microfabricated devices could become microsurgical tools, fluid pumps, meters and valves. Three basic technologies are used to create microfabricated devices 50 microns in diameter (about half the diameter of a human hair). Silicone microstructures are created by surface machining, bulk rnicromachining, and Lithographic Galvanoforming Abformung (LIGA). Garcia concluded his talk by showing a videotape of a high speed micromotor.
Marvin Vestal spoke on recent developments in matrix-assisted laser desorption time-of-flight mass spectrometry, which have allowed the implementation of automation to facilitate faster throughput of samples. Mass spectral data are automatically acquired and processed from samples that have been automatically loaded onto sample plates. Sequence ladders of proteins and oligonucleotides can be determined from data obtained on samples enzymatically digested on the plates. This acquired data can be processed for rapid structure determination using an automated routine.
Medical robotics will revolutionize the speed at which human surgery can be performed. In the future a single surgeon will be assisted by a team of robots that will perform minirnally invasive surgery with a better outcome to the patient. However, novel surgical tools must be developed to assist the surgeon with the many tasks associated with even the simplest of human surgeries. Anita Flynn has developed a high torque (100–300 rpm, 1 second spin -up, 41 pnewtons of force) low mass motor which can form the basis of many automated surgical tools. Using ultrasonic principals, she developed an ultrasonic motor which did not require the typical gears associated with conventional high torque motors. Flynn demonstrated an intestine crawling robot in its rudimentary form which could perform endoscopies in the future.
Routine laparoscopics (examination of the intestinal cavity using video endoscopes) require the assistance of a nurse or surgical assistant to hold and guide the video system which provides the surgeon with visual reference for manipulations. Both of the surgeons hands are typically occupied with surgical tools. Aesop, a novel video tool-holding robot, can be guided and positioned using voice commands. Yulun Wang developed Aesop to enable solosurgery (unassisted surgery by a single physician), robotic-enhanced surgery (allow micro positioning unattainable by a human operator), manipulation enhancement (the extra pair of hands necessary to perform difficult procedures), and telesurgery (surgery performed at remote sites by a surgeon at a central location).
Teaching and guiding robots can be a frustrating experience since they don't yet have the dexterity or spacial sense of humans. The creation of robotic autonomy is the subject of Polly Poole's research into diectic teleassistance (from the Greek word for “to point”). Pook demonstrated that robots could be taught to perform complex tasks from simplified sign language which provides a simple link to stored robotic movements. For example, if a robot has already been trained to open a door handle, the operator does not have to demonstrate the entire procedure to the robot each time. Instead, the operator simply indicates to the robot the desire to have the door opened. This action initiates a series of stored movements which have been previously shown to have the desired outcome. This method overcomes sensor limitations, real time constraints of feedback mechanisms, and eliminates the need for tight servo-control.
1996 ICAR Sponsors
ICAR '96 wishes to thank the following sponsors that supply quality automation technologies and have a commitment to education in this rapidly evolving field:
Abbott Laboratories
AutoLab Systems-A Division of MDS Health Group Ltd.
Beckman
Becton Dickinson Vacutainer Systems
Boehringer Mannheim
Ciba Corning
Coulter Corporation
DuPont Medical Products
Johnson & Johnson Clinical Diagnostics
Hitachi
Medical Automation Systems
Olympus Corporation
Packard Instruments
Sysmex
The program concluded with one of the most exciting technologies in the automation field, wireless communication. Fred Velazquez (Boston University Medical Center, Boston, MA) described the many facets of wireless communication. The goal today is to be able to transmit the maximum amount of data in the minimum time over the widest geographical area. Many options for communication are becoming available as a result of cellular telephone networks. For example, cellular digital packet data is an error correcting method to send bursts of data over the cellular network for only pennies per kilobyte. Spread spectrum wireless is less prone to interference but only has a 500-foot range. The applications of wireless communication has to be married with the best transmission method to yield the desired communications network Fred concluded with a description of a medical communication system developed in Boston with the help of Boston Medical Center.
1995 AWARD RECIPIENTS
The Hewlett Packard Award, given for outstanding strategic research in automated sample preparation, was presented to Don Hagen. Hagen was recognized for his seminal research on particulate loaded membranes, which form the basis for the 3M company's Empore™ solid phase extraction discs.
Don Hagan receives the Hewlett Packard award, from Conference Chairman Robin Felder, who was standing in for Hewlett Packard's Brian Holden.
Particulate loaded membranes, which have enormous binding capacity, can be utilized to isolate selected classes of analytes from mixtures in several microliters to many liters. This technology has already been incorporated into several EPA methods. Hagen's work (with partner Craig Markell) has resulted in a number of technical publications, presentations, and patents.
Hagen is a corporate scientist at the 3M Company in Saint Paul, Minnesota. He received his BS degree from the University of Wisconsin, Madison in 1957, and his MS degree from Oklahoma State University in 1961. He has been with 3M since 1963. His previous awards include the 3M Analytical Systems Development Award (1977), the Minnesota Chromatography Forum L.S. Palmer Award for Outstanding Contributions in Chromatography (1982), the 3M Calton Society (1986), and the American Chemical Society Award for Contributions to Chemistry (1994).
Kevin Bennet was chosen to be the recipient of the Boehringer Mannheim Award for pioneering and original research in the general field of automation. Bennet developed the analytical framework for evaluating automation opportunities at the Mayo Clinic, utilizing computer simulation and financial modeling. He also developed a laboratory automation plan for the new Mayo Laboratory Network facilities. Through these studies he designed a continuous automated clinical centrifuge which has been licensed and commercialized.
Bennet currently holds several appointments within the Mayo Clinic and Mayo Foundation. In Systems and Procedures he is an internal consultant involved in laboratory automation and the operation of the medical practice. In Strategic Alliances he develops relationships in research and product development with organizations outside of Mayo. He also holds an appointment as Instructor in Tumor Biology at the Mayo Medical School. Bennet received his BS in Chemical Engineering from Massachusetts Institute of Technology and his MBA from Harvard Business School. He has published on a variety of subjects and holds a patent involving the production of high-resolution ceramic electronic packaging.
Kevin Bennet, recipient of the Boehringer Mannheim award for original research in the field of automation.
ORAL AND POSTER PRESENTATIONS AWARDED AT ICAR '96
Cargill
Khoulif
An award was presented at ICAR '96 by Scitec-TNO to John Cargill for his oral presentation, “Automated Combinatorial Chemistry on Solid Phase.” Cargill is a principal scientist with Ontogen Corporation, where he has been responsible for developing hardware and software systems for the automation of combinatorial chemistry. He joined Ontogen in 1993 as head of the laboratory automation group; prior to that he was manager of Research Computing at the Ontario Cancer Institute in Toronto.
The prize for best poster was given to Zidene Khoulif of the Conservatoire des Arts et Metiers, Paris, for his poster, “Robotization of the On-Line Analysis: Application to the Flexibilizer Fabrication Process.” Khoulif's work at CNAM involves the implementation of a robotic system for performing organic synthesis at the laboratory scale.
