Studyledge

Automated Systems, Robotics and Artificial Intelligence: Question 4

Syllabus 6.3

Structured 9 marks

The technology club at a school has an expert system installed on the computer that controls its 3D printer. When a print fails partway through, a student technician runs the expert system: it asks a series of on-screen questions about what the failed print looked like, and then displays a diagnosis of the likely fault together with a suggested fix.

(a) State the four components of an expert system. [4]

(b) For the 3D-printer expert system described above, explain how the interface, the knowledge base and the inference engine work together to reach a diagnosis. [3]

(c) Before the expert system was installed, the club used a single printed page of fixed yes/no troubleshooting questions instead. Explain one way in which using the expert system, rather than this fixed printed page, benefits the club. [2]

Show worked solution Hide worked solution

Worked solution

Part (a): The four components of an expert system

An expert system is made up of:

  1. A knowledge base, the stored facts about the subject (here, common 3D-printer faults and the symptoms linked to each one).
  2. A rule base, the set of rules built from that knowledge, used to link symptoms to conclusions.
  3. An inference engine. The part that applies the rules in the rule base to the data entered by the user to reach a conclusion.
  4. An interface, the part that lets the user enter information and that displays the resulting diagnosis.

Part (b): How the interface, knowledge base and inference engine work together

  • The technician enters the symptoms of the failed print (for example, answering “yes” or “no” to “did the first layer stick to the bed?”) through the interface.
  • The inference engine takes these answers and works through the rules in the rule base, checking them against the facts stored in the knowledge base about which faults produce which symptoms. For example, that a nozzle set too far from the bed is linked to the first layer failing to stick.
  • By eliminating rules that do not fit the answers given, the inference engine narrows the possibilities down to a diagnosis and a suggested fix.
  • This result is sent back through the interface to be displayed to the technician.

Part (c): Benefit over the fixed printed page

The expert system’s inference engine can combine several of the technician’s answers together and choose which question to ask next based on the previous answer, working through many different combinations of symptoms to reach a specific diagnosis. A single printed page of fixed yes/no questions cannot adapt in this way. It can only offer the same limited set of prewritten explanations regardless of how the answers combine, so it is more likely to miss faults with unusual or combined symptoms that the expert system’s reasoning can identify.

Final answers

  • (a) Knowledge base, rule base, inference engine, interface.
  • (b) Interface takes in the technician’s answers → inference engine applies the rule base to the knowledge base → diagnosis is displayed back through the interface.
  • (c) The inference engine can combine answers and branch to different follow-up questions to reach a specific diagnosis, which a fixed printed checklist cannot do.