Is Model Based Reasoning Right for You?

If you’ve never heard of Model Based Reasoning (MBR), your first thought might be: isn’t all reasoning “model based”? And on learning that MBR is a form of artificial intelligence, your second thought might be: who would trust AI to do a human job like “reasoning”?

MBR clearly has to cope with some unfortunate and confusing labels. But that hasn’t stopped service organizations around the world from using it, profiting from it, and vouching for it in impressive numbers. Every aftermarket service professional should know what it is, and how they stand to benefit.

Put simply, the desktop of a human mind is small and rickety. We may master a large body of knowledge, and may be able to recall most of its parts, given enough time and prompting. But given a few ambiguous symptoms, only the most capable and well-trained humans can summon all the relevant facts and principles, and make a sound inference in a short time. That’s precisely why Sherlock Holmes, and characters like him, seem superhuman. They think on a sprawling mental desktop, where information can be gathered, sorted, and assembled with astonishing speed.

Instead of a desktop, the rest of us have something like a writing slab in an old lecture hall. It’s barely large enough for a single notebook, and barely stable enough to take notes on. It can be expanded, but only with rare talent or expensive training. Otherwise, when making a diagnosis, we’re forced to scrutinize one possibility at a time. This is what leads to “linear reasoning” – an unreliable and slow approach, where the possible causes of each symptom are individually considered, usually in an arbitrary order. Service providers who rely on linear reasoning have to tolerate longer Mean Time to Repair, and absorb larger losses from device downtime.

Model Based Reasoning is a popular shortcut. With MBR, service experts can codify their knowledge, so that a program can apply all of the facts, in unison, to a set of symptoms. This “brute force” approach means a lot of work for the computer, but luckily that’s what computer are for, and so far they haven’t complained. With this “non-linear” technique, technicians can overcome the limitations of human memory and attention, quickly computing what the present symptoms mean, what the absent symptoms mean, and what possible malfunctions are left. MBR does the work of a large team of human workers through a single program, sifting through the possibility space, and leaving a small set of suspects for the technician to investigate.

Needless to say, MBR’s applications are varied, and its benefits widely recognized. MBR has proven invaluable in Guided Troubleshooting, Diagnose before Dispatch service, Remote Diagnosis, and Design for Service testing. Nearly all companies surveyed said that their use of MBR had reduced diagnostic time and improved service productivity. For aftermarket service professionals who deal with complex product failures, devote inordinate time to troubleshooting, have little service history on which to base diagnoses, work with frequently updated products, or conduct remote or automated diagnoses, MBR platforms may be just the ticket. To help you find an MBR platform for your aftermarket service needs, there’s always Blumberg Advisory Group. See our latest whitepaper on MBR, or schedule an appointment to decide what MBR solution is right for you.


Please note: I reserve the right to delete comments that are offensive or off-topic.

Leave a Reply

Your email address will not be published. Required fields are marked *