October 30, 2017

Transforming 6 Operations across 15 countries for a $10 Billion Process Chemicals Giant [Use Cases]


Cerebra Quality- IoT Intelligence for Manufacturing Plants
IDG

A lot has been talked about the hype around #Industrial IoT, viable business case, a juicy ROI for IoT platforms. I wanted to share this very proud moment where #Flutura's Cerebra deployment which started as a small app back in 2015 is now the central nervous system driving plant operations worldwide for a Fortune 500 process chemical major.
The customer's global operations and quality teams called for a checkpoint review last week to assess the impact of Cerebra on its operations and looking at new avenues to leverage it for its operations.
We never imagined the number of ways Cerebra could impact day to day operations as it is today. Here are a few areas where Cerebra has triggered concrete actions to transform operations.
1.  Cerebra drives product quality benchmarking - Before Cerebra, every product manufactured had its own quality measurements and was difficult to assess, benchmark and compare the quality across manufacturing batches, across production lines, and across factories to take corrective actions. Cerebra introduced a scientific way to measure product quality in a homogenous manner enabling measurement of performance at all levels. At a single instance, the product quality can be assessed in a simple and intuitive way. The entire plant management team now reviews operations through Cerebra. Old, paper-based review processes have been discontinued.
2. Cerebra learns complex interactions - Before Cerebra it was impossible to attribute causal factors for suboptimal batches due to the multitude and complex nature of data points available. Cerebra is now constantly learning the ever-changing complex interactions and inter-relationships between human, machine, environment, processes, and outcomes throughout the manufacturing process and helping operations and quality teams to intervene at the right time.
3. Cerebra is “THE” root cause analysis system - Before Cerebra, root cause analysis was manual, highly time-consuming and many a time impossible to conduct due to the sheer complexity of the process. Cerebra is now able to isolate the needle in the haystack with ease; one of the manufacturing plants was able to isolate an issue related to the design of the blades used in the industrial mixer to be the cause for variance in viscosity of the manufactured product. The production schedule was changed in order to circumvent this constraint
4. Cerebra pinpoints critical manufacturing step changes - Before Cerebra, macro deviations or step change in operations were difficult to detect. Cerebra is now helping detect macro shifts in processes; a manufacturing plant was able to detect a change in product quality due to a change they made to a supplier of a critical raw material several months before. Cerebra now helps review Supplier performance.
5. Cerebra detects unforeseen patterns - Before Cerebra, repetitive & seasonal patterns were difficult to detect. Cerebra is now helping unearth unforeseen changes to environment leading to undesirable impact; Cerebra detected a peculiar pattern related to variance in quality from a particular shift, on the investigation, the training process was relooked into and operators are now being trained on the new process.
6. Cerebra optimizes Standard Operating Procedures - Before Cerebra, process effectiveness and continuous improvement to the standard operating procedure were subjective and reactive. Cerebra is now helping monitor compliance to standard operating procedures, the effectiveness of standard operating procedures and helping question status quo; Cerebra helped narrow down more-than-a-decade-old quality threshold established for a specific product's BOM and in turn, helped change the product development formulation.
Cerebra IIoT intelligence platform was initially deployed to gain basic visibility into product quality deviations on the manufacturing line; has now
1. Transformed Production Planning
2. Transformed Machine Resource Planning
3. Transformed Supplier Performance Assessment
4. Transformed Product Development
5. Transformed Manufacturing Process Optimization, and 
6. Transformed Operator Training.
Many plans and blueprinting are underway to leverage Cerebra in multiple areas; next being Supply Chain Optimization. 
Cerebra is serving as THE operating system on which Man-Machine interactions are being orchestrated. It has become the Central Nervous System for plant operations.
Can't wait to see a day when Cerebra supports autonomous operations.
It was an amazing moment to see the team rejoice at the impact they had made to our customer's business. Great job team. Look forward to many more transformations

October 5, 2017

Cerebra Feature Series #2 - Discovering Unknown Unknowns [Use cases]

"I have lots of data and I don't know what to do with it"
"We collect a lot of data but not sure whether it is accurate or has any value which can be unlocked"
"We have conducted a lot of experiments and proof of concepts but very few of them have resulted in big operational or financial impacts"
These are the most common experiences we hear in our conversations with customers. These are definitely not trivial questions. We call it the Data-to-$-Chasm.The questions themselves not only relate to unidentified problems to solve but also includes unidentified answers to complex questions which needs macro and micro synthesis of large volumes of data.
Now, does this exercise of finding problems to solve, or finding answers to questions which needs large breadth of data worth the effort ? What if these experiments don't yield a BIG return on effort or investment. The short answer is "YES this is worth the effort"
We have been working on solving this problem for sometime now and has resulted in a core product suite within Cerebra called Pre-Diagnostics to aid engineers' maieutic enquiry process.
In the engineering world, first principles based interrogation techniques are still widely adopted, however when Physics + Chemistry + Ambient condition + Operating Procedures + Human behavior is involved, currently first principles based modeling is complex and sometimes unattainable. You never know there might be a day when the "real world" can be expressed with a universal equation !
For the time being let us stick to Pre-Diagnostics, These are some of the "Never before seen insights" our customers discovered using Cerebra Pre-Diagnostics, which enabled operational process changes and modifications to their industrial equipment ,
  1. A leading industrial glove manufacturer discovered unusual changes to sensor behavior affecting product quality with Cerebra's Swarm Algorithm. Pollen pollution in the shop floor was ascertained as the root cause for this issue and extensive changes were done to set right the ambient operating conditions for the plant. Continuous monitoring of ambient conditions has now been institutionalized
  2. A process chemicals manufacturer discovered deviations to the composition of the end product while it was manufactured with a particular industrial mixer. This was a decade old problem which had gone unnoticed. On further investigation it was ascertained that the type of blade used in the industrial mixer was leading to deviations in viscosity of the product. The machine resource allocation process has now been changed to add constraints related to product - machine mix
  3. A gas processing plant had frequent failures on its valves resulting in periodic unplanned downtimes. The discovery process lead to surfacing insights related to non synchronized feedback loop from the control systems to the valves leading to friction. The plant is replacing its control systems architecture to prevent issues
  4. A leading mining company discovered failure patterns on their milling machines when certain load thresholds were breached consistently over a period of time. New alarms were designed and put in place to detect progressive failure paths
Not long ago people could imagine only white swans, because white swans were all they had ever seen. And so people predicted that every next swan they would see would be white. The discovery of black swans shattered this prediction. The black swan is a metaphor for the uselessness of predictions that are based on earlier experiences, in the presence of unknown unknowns -Taleb,Nassim
Cerebra Pre-Diagnostics helps in identifying problem statements worth solving and use cases worth operationalizing in a short time
Watch out for the next feature in Cerebra Feature Series.

October 3, 2017

Cerebra Feature Series #1 - Digital Twin in 48 Hrs [ Case Study ]



Flutura's team achieved the improbable when they enabled a massive oil field equipment used in early production stages of crude oil extraction to be connected in 48hrs and turned on Cerebra's diagnostics engine to look at operational insights and anomalies of this critical equipment from anywhere in the world.Prior to this, a lot of manual monitoring of the flowrates and other vital signals used to happen.
The machine by itself is a complex integration of Desander, Separation Units, Outlets, Valves and other machine units. The machine is subjected to high pressure of up-to 9000 psi which is the norm in any Oil field equipment. Data from close to 500 sensor tags across the machine units are streamed every second. Cerebra ingests this data with very high reliability and broadcasts critical signals to key stakeholders in real time.
Cerebra's universal machine model helps model any machinery and equipment with ease.
Watch out for the next feature in Cerebra Feature Series.