ALBERT

Notes: This paper describes several prototypical applications of neural network technology to engineering problems. The applications were developed by the authors as part of a graduate-level course taught at the University of Illinois at Urbana-Champaign by the first author (now at Carnegie Mellon University). The applications are: an adaptive controller for building thermal mass storage; an adaptive controller for a combine harvester; an interpretation system for non-destructive evaluation of masonry walls; a machining feature recognition system for use in process planning; an image classification system for classifying land coverage from satellite or high-altitude images; and a system for designing the pumping strategy for contaminated groundwater remediation. These applications are representative of many of the engineering problems for which neural networks are applicable: adaptive control, feature recognition, and design.

Notes: Abstract Near-infrared chemiluminescent emission from NeMu*, the analogue of the Rydberg molecule NeH, has been observed in Ne, Ar, and Ne/Ar gas mixtures. Three temporally distinct features were observed: First, a large sharp emission peak at time zero, observed in all gases (Ne, He, N2, Ar), is assigned to scintillation light during muon thermalization, probably caused by spur electrons. Second, a lowintensity broad region observed in all gases is attributed to e+ from muon decay. Finally, NeMu in 1–6 atm Ne with 0.1–2 torr Ar appeared as a high intensitydelayed emission, whose width and intensity depended linearly on the Ar concentration. Its wavelength spectrum from 680–960 nm was measured. Although questions remain as to how NeMu* is formed, the precursor is likely Neμ +. Possible electron donors include metastable Ar* (3 P 2 or3 P 0) and long-lived free (spur) electrons.

Notes: Abstract In order to ensure the stability of protein pharmaceuticals, human serum albumin (HSA) is often added as an excipient, frequently in large excess. This makes chromatographic analysis of the stability of the active protein difficult. In the case of interleukin-4 (IL-4), separation from HSA can be achieved to some degree by size exclusion chromatography, but some HSA co-elutes with the IL-4. Hydrophobic ion pairing provides a method for selective precipitation of IL-4 from HSA. Hydrophobic ion pairing involves the electrostatic interaction of ionic detergents with oppositely charged polypeptides. Even when HSA is present in fifty-fold excess (w/w), the resulting precipitate contains greater than 70% of the IL-4. Selective precipitation with SDS produces enhancements in IL-4 over HSA of more than 2000-fold. This approach permits subsequent facile analysis of IL-4 by conventional reverse phase HPLC.

Notes: Abstract The Solid State Imaging (SSI) experiment on the Galileo Orbiter spacecraft utilizes a high-resolution (1500 mm focal length) television camera with an 800 × 800 pixel virtual-phase, charge-coupled detector. It is designed to return images of Jupiter and its satellites that are characterized by a combination of sensitivity levels, spatial resolution, geometric fiedelity, and spectral range unmatched by imaging data obtained previously. The spectral range extends from approximately 375 to 1100 nm and only in the near ultra-violet region (∼ 350 nm) is the spectral coverage reduced from previous missions. The camera is approximately 100 times more sensitive than those used in the Voyager mission, and, because of the nature of the satellite encounters, will produce images with approximately 100 times the ground resolution (i.e., ∼ 50 m lp-1) on the Galilean satellites. We describe aspects of the detector including its sensitivity to energetic particle radiation and how the requirements for a large full-well capacity and long-term stability in operating voltages led to the choice of the virtual phase chip. The F/8.5 camera system can reach point sources of V(mag) ∼ 11 with S/N ∼ 10 and extended sources with surface brightness as low as 20 kR in its highest gain state and longest exposure mode. We describe the performance of the system as determined by ground calibration and the improvements that have been made to the telescope (same basic catadioptric design that was used in Mariner 10 and the Voyager high-resolution cameras) to reduce the scattered light reaching the detector. The images are linearly digitized 8-bits deep and, after flat-fielding, are cosmetically clean. Information ‘preserving’ and ‘non-preserving’ on-board data compression capabilities are outlined. A special “summation” mode, designed for use deep in the Jovian radiation belts, near Io, is also described. The detector is ‘preflashed’ before each exposure to ensure the photometric linearity. The dynamic range is spread over 3 gain states and an exposure range from 4.17 ms to 51.2 s. A low-level of radial, third-order, geometric distortion has been measured in the raw images that is entirely due to the optical design. The distortion is of the pincushion type and amounts to about 1.2 pixels in the corners of the images. It is expected to be very stable. We discuss the measurement objectives of the SSI experiment in the Jupiter system and emphasize their relationships to those of other experiments in the Galileo project. We outline objectives for Jupiter atmospheric science, noting the relationship of SSI data to that to be returned by experiments on the atmospheric entry Probe. We also outline SSI objectives for satellite surfaces, ring structure, and ‘darkside’ (e.g., aurorae, lightning, etc.) experiments. Proposed cruise measurement objectives that relate to encounters at Venus, Moon, Earth, Gaspra, and, possibly, Ida are also briefly outlined. The article concludes with a description of a ‘fully distributed’ data analysis system (HIIPS) that SSI team members intend to use at their home institutions. We also list the nature of systematic data products that will become available to the scientific community. Finally, we append a short ‘historical’ note outlining the responsibilities and roles of institutions and individuals that have been involved in the 14 year development of the SSI experiment so far.

Notes: Summary Recombinant interleukin-2 (rIL-2) modified with monomethoxypolyethylene glycol (PEG IL-2) was utilized in patients with metastatic renal cell carcinoma in two separate multi-institutional trials. PEG IL-2 was administered as an I.V. bolus days 1, 8, 15, and 22 with cycles repeated every six weeks. The two trials employed different dose levels: A) 20x106 I.U./m2 day 1 followed by 12x106 I.U./m2 days 8, 15, 22; and B) 12x106 I.U./m2 days 1, 8, 15, 22. Thirty-five patients were entered and 31 were evaluable for response (A−15/18, B−16/17). Two of 31 patients had partial responses. Median therapy duration was four weeks (range 1–15), and dose reduction for grade III or IV toxicity was required in 14/35 patients (A-6/18, B-8/17). Toxicity (≥ grade III) seen included: hypotension 51%, dyspnea 17%, seizures 6%, and mental status changes 11%. No differences in response or toxicity between the two schedules were noted. Hematologic changes included lymphocytosis and eosinophilia in the majority of patients. PEG IL-2 given once weekly has significant toxicity, and may produce tumor regression in patients with renal cell carinoma.

Notes: Abstract In an effort to provide a better understanding of the sources of petroleum hydrocarbons in urban runoff, petroleum products as well as probable source materials were analyzed for hydrocarbons and trace metals and then compared to urban runoff samples from four different land use areas. The petroleum products considered were No. 2 and No. 6 fuel oils, used and virgin crankcase oils, and gasoline. Source materials included street dust, roadside soil, roadside vegetation, and atmospheric fallout; and the land use categories were commercial, residential, interstate highway, and industrial areas. The samples were compared on the basis of hydrocarbon (saturated and aromatic) and trace metal (Cd, Pb, Mn, Cu, Fe, Zn) loadings and distributions. Results indicated that the hydrocarbon content in runoff at all four land use sites originated primarily from used crankcase oil, with a small amount of No. 2 fuel oil detected at the industrial site. Only a small portion of the crankcase oil component came from the sources surveyed, and the majority of this oil probably came from: (1) oil drops within the driving lanes on the road surfaces or deposits in parking areas, and/or (2) direct dumping of waste crankcase oil down storm drains.

Notes: Abstract A direct causal relationship was demonstrated between soil temperature and insect ovipositional propensity. When ovipositional substrates (soils) at 5, 15, 22, 30, 35, and 40°C were presented in multiple treatment (‘choice’) tests with air temperature at 15 or 22°C, onion flies, Delia antiqua (Meigen) (Diptera: Anthomyiidae), laid the most eggs in the 22°C substrate. Only 50 eggs were laid when air temperature was increased to 30°C, as compared to 454 and 1128 eggs at 22 and 15°C, respectively. Thus, an air temperature of 30°C appears to be near the upper limit of onion fly ovipositional activity. The numbers of flies observed (counts taken every 15 min) on substrates ranging from 15 to 40°C were not significantly different. Reduced alightment/arrestment does not explain reduced oviposition on the warmer substrates; however, it may partly explain reduced oviposition on 5°C substrates. The range of substrate temperatures facilitating substantial oviposition was narrower than that eliciting alightment/arrestment on the substrate. The ca. 20°C ovipositional optimum corresponds well with temperatures favoring egg survival and development.