Alex d’Agapeyeff, working with the British Computer Society Expert Systems Group and under the auspices of the Alvey Programme, has just completed his second survey of implementation of Expert Systems in the UK. And the message is that the UK is in much better shape three years after his first survey, has almost caught up with the US and Japan in terms of systems implemented per company, and is on the brink of even greater success. But he claims that UK bosses aren’t interested in the systems, and departmental managers are being made to carry the flag. Where their US counterparts thought of expert systems as part and parcel of strategic planning, in the UK the highest levels of management ignore them. The other problem holding back UK progress is secrecy. The more successful an implementation is, the more it is viewed as company confidential or worse. D’Agapeyeff has even come across one company that had two project managers, each building expert systems, yet neither had any idea what the other was doing. Learning was duplicated, as were resources, and d’Agapeyeff believes that the veil of secrecy prevents necessary constructive criticism and sharing of ideas, and slows the whole process. The largest teams have achieved between eight and 12 live applications, but there is a larger band of firms with just two to five applications running. Once the first system is installed, says d’Agapeyeff, the benefits are so obvious that the company becomes committed to implementing more, which makes it crucial that the first one works. There were companies that had given up with expert systems because the person in charge of the first development had moved office, and other managers just anticipated that their development was either costly or not worth having, which he says is not true. The report was compiled largely based on the experience of 15 companies, computer firms, consultancies and users alike.