Big data analytics is a bigger topic today and use cases are arising in many areas of business. These range from the traditional CEP areas of financial trading to new customer engagement models such as geo-located coupons. In an Internet of Things (IoT) world, machine to machine (M2M) solutions can produce big data requirements. These, ‘edge devices’ can post vast volumes of data into the cloud applications that service them. Manufacturing is an area, under-served by BPM and there certainly are numerous industrial M2M and big data use cases. At Bosch, we believe one of the most important and fruitful areas are predictive maintenance (PrM).
Factory floors frequently have expensive and complicated machinery that is critical to their manufacturing competitiveness. The unplanned loss of even a few Hours can result in a great financial loss. . Many companies are turning to the big data, analytics topic of predictive maintenance to optimize these resources. These methods analyze the condition of in-service equipment to project optimal maintenance measures. Analytics for this fields encompasses a broad range of mathematical disciplines, for instance vibrational analysis can use the occurrence of even a small 'out of round' sensor reading to predict the failure of a machine. When we start to aggregate a number of machines, one can apply data heuristics discover even more predictive factors. Even minor irregularities and latent failure patterns are uncovered and aligned across the resource pool.
It is not enough to have the mathematical, material science and engineering capability to create a predictive maintenance solution—— there needs digital infrastructure to support this. That infrastructure must also be capable of hosting M2M solutions. Bosch software already has an example of this in the Service Portal.
The service portal uses A combination of BPM+ and the output of predictive analytics to generate the appropriate response.
On June 6, 2013 Bosch Software will hold a webinar on their approach to PrM.
I will be speaking along with my colleagues Karsten Koenigstein and Christina Gruen.