My colleague’s article on Edge Intelligence is well timed. Economic and environmental pressures are forcing companies, and indeed entire industries, to find ever more innovative solutions to optimize their processes and reduce down time. The shift from traditional manufacturing processes towards Smart Factories or Industry 4.0 requires a combination of smart sensors, wireless and wired communications, cloud data and analytics and perhaps most importantly, innovative and pioneering leaders.
At Cambridge Consultants we specialize in industrial sensing technology, smart algorithms, machine learning, pattern recognition, instrumentation and big data analytics and have implemented these skill-sets in a number of sectors from Oil & Gas through to Consumer. These have included industrial products and systems such as the cloud based identification and positioning of underground cabling with IPEG, warehouse automation solutions with Ocado, condition monitoring with DropTag Drive through to consumer technology developments such as the analytics of a baseball bat swing with the SparTag smart sensor.
Our industrial product development team have backgrounds in large scale industrial process plants and factories and are working on the technology transfer of our smart sensor knowledge to some of the world’s largest industries such as: mineral processing; sugar and bio-energy; water processing; cement production; steel; aluminium and other metals production; manufacturing; mining; and energy generation from coal, renewables or nuclear.
All of these industries use a large amount of process equipment, but in fact many of these equipment types are quite similar regardless of the industry: a bag filter in an aluminium smelter is very similar to that used in a power plant or food manufacturer and the same applies to many pumps, belt and screw conveyors, vibrating feeders, crushers, ball mills, bucket elevators, etc.
But what does this mean practically? We are looking at more and more innovative ways of combining low cost, smart sensors and instruments like the accelerometers used in SparTag or the load cells used in SmartDrawer with smart algorithms and software to extract meaningful and actionable data that can be used for process optimization and preventative maintenance in these high volume, tight margin industries. We are looking at the extraction and analysis of information that wasn’t previously obvious or available. Machine learning, event detection and pattern recognition to: anticipate bearing failure in a turbine; a water pump drawing air through the system; a mixer with an incorrect balance of input materials, or possibly even materials with variable properties (moisture content, particle size) or a mill vibrating at different frequencies based on the material properties at the time. All of this embedded in a smart sensor that is capable of analyzing critical information and transmitting only this relevant data over the network.
The earlier, or further up-stream, this information or these events can be identified the better the process can be optimized or preventative maintenance carried out, reducing downtime, cost and increasing yield.
One thing is certain, the industrial landscape is changing, with businesses looking to understand and embrace a diverse range of new technologies to maintain or gain a competitive advantage.