The computerized niche in mining research continues to grow. Computer modeling in mining began with ventilation and quickly expanded into pillar design, subsidence, and hydrology. One interesting program is MULSIM/BM from the US Bureau of Mines. It is a rock stress program designed to assist engineers in the layout of mine plans. The program calculates stresses and ground deformations resulting from mining. Special features include the ability to include up to 26 material property sets, subdivide to "fine mesh" portions, and to specify extraction ratios. The designers say the program should be "used discriminately in combination with field verification data." That is an honest caution statement that every promoter of a computer model should be using. Another Bureau program that has ready application is one that forecasts subsidence over longwall panels. In some areas, enough modeling has been done for someone to undertake the task of comparing programs. Predicting water inflows into operating coal mines is an active CONOCO research area. And researchers in the Library, PA coal research division have undertaken this difficult task with positive results. Computer applications dominated the second conference on innovative mining systems, held in October at Pennsylvania State University. Here, the emphasis was more on computer control of mining equipment. Drilling is the most down-to-earth application discussed, and computerized drills have just recently come into use. Another recent mining application is control of longwall shearers. However, it is unlikely that these rather straightforward uses mean that widespread computer control of mining equipment is just over the horizon. Future advances may be in better-funded areas -related to mining, such as emplacement of nuclear waste. So-called "expert systems" represent another interesting computer application, where the Computer is proported to have some characteristics of human thinking. Applications are visualized in mine ventilation, equipment control, and problem diagnosis. But, aside from a system called Prospector developed a decade ago at Stanford, virtually every mining application remains in the talking stage. One interesting system that may soon see application is called MICA, a joint development of the Bureau and the University of Missouri. It acts as a consultant to provide advice on mica beneficiation. In the area of coal use, there were some interesting developments in microbubble flotation, which offers better selectivity for smaller particle sizes. Research has demonstrated that, in addition to coal mineral characteristics and the reagents used, air bubble size plays a vital role in flotation. Bechtel National has teamed up with Energy International to use a unique flotation cell developed by Bergbau Forshung in Germany. The Bergbau flotation cell is a conical vessel using a pre-aerated microbubble slurry. It has maximum bubble loading with even the finest particles. Also, Lurgi Corp. reported on its operating experience on its dry/semi-dry circulating fluid bed process for flue gas desulfurization. Some advantages claimed are dry reagent feeding, no slurry atomizer, higher solid residence time, and ability to reduce any S02 amount from very high sulfur coals to legislated levels.