A recap of yesterday's webcast on driving productivity and profit through energy management technology, and the top questions answered.
January 19, LNS Research hosted the webcast, “Modern MES: Are Companies Finally Ready for Cloud Options?” The broadcast featured Bloom Energy’s CTO, Venkat Venkataraman who presented their findings and experiences with Cloud MES. Industrial companies and manufacturers tuned in to get an accurate understanding about the realities of today’s options. Mr. Venkataraman explained how Bloom selected the right MES solution for their new plant, then rolled it out to all the other facilities in its network.
5 Top Questions
Q1: “What does the migration look like and do they have to work on this new platform? As you can imagine there is a lot of legacy and move can be tough. What are some of the things we can do get ready?”
A1: We do not see that much difference in going to a Cloud solution rather than a more traditional MOM system. In the Cloud scenario, starting small is probably easier (and certainly less expensive).
Q2: “How effective the system for decentralized manufacturing facility?”
A2: Very! Roll out from a single multi-tenant solution will be less painful than plant-by-plant installations.
Q3: “What does a company need to have regarding people, process, and technology before considering a Cloud-based MES system?”
A3: Great question that we will be addressing in upcoming research – all three are vital to success.
Q4: “Should customers be concerned about multi-tenant SaaS solutions? Should they insist on a partitioned, isolated, single-tenant solution?”
A4: You certainly should be concerned, but not fearful. Make sure that your supplier has the right experience in delivering a secure, performant and available public Cloud solution.
Q5: “Where are the production data actually stored? According to your experience, which percentage of manufacturing companies are reluctant to "lose " control of their data?”
A5: Production data is stored in the Cloud and available to those who have permission to access it. Many manufacturers are, initially, reluctant (this is a euphemism) to give up “control” of their data. When business leaders start to take industrial transformation seriously, wider access to production data becomes a business priority, and wider access will be mandated. At this point, moving data into the Cloud will help a lot in securing it and, at the same time, making it more broadly available.
We received a couple of questions about standardizing solutions across plants and access to information from anywhere. For me, the issue of standardization is one of the greatest benefits of a single solution delivered from the Cloud to all facilities. As you roll out new facilities, the normal mode of working will be to use the standard template that is already live for other plants. Changes will be required, but the goal will always to minimize them, and certainly to avoid any changes that disrupt existing plants.
One of the big advantages of a single Cloud instance for all plants is standardized visualization and reporting. While new reports might be required as time goes on, they will always be developed to apply to all lines and facilities. This will deliver access to any information for everyone regardless of their current location.
Needless to say, we had some questions regarding speed, latency, and availability. One attendee asked: “Consider a scenario of having 40+ plants for which MES/MOM is on Cloud. If the scheduling of all these plants is done from a central location will latency effect be a stumbling block?” I think the answer to this particular question is a simple “not if you design it right.” In fact, scheduling across all 40 plants will bring a closer to optimal schedule than a simpler solution. Of course, you will not want to reschedule all 40 plants every time a small change in a plant requires some local rescheduling. Having capabilities to adjust a schedule and report what has been changed can be built into the planning and scheduling system. As we mentioned in the webcast, going to the Cloud does not mean that pragmatic solutions to real world problems will be forgotten: this is a classic case of keeping it simple to ensure that the MOM system works for rather than against its users.
Data Volume and Speed
A few questions on data volume and how MOM in the Cloud handles it included:
- In a high-volume production environment, like cell phone manufacturing, there are gigabytes of data that need to be stored daily. Do you have an example of how you used Cloud in an environment like that?
- How much data of manufacturing is actually stored in a Cloud? ( e.g. past three months of production data or 6 months or 1 year).
- What is the amount of network traffic required for this Iot MES?
- What is the volume of data that can be collected from the shop floor? 1 reading per second? One reading per millisecond?
One of the huge benefits of using a public Cloud service is scalability. When designing the solution, you do not have to decide on storage capacity nor do you have to pay for capacity that you are not yet using. SaaS is a pay as you go and use model, both for license and storage. This model also helps when considering how long you will keep your data; not only can you change your mind but you can change your storage requirements as business needs change. For example, you might start off keeping manufacturing data of delivered products for, say six months; if a new customer requires that you keep the information for three years, the only impact is your monthly subscription charges for more data use. Sanmina has used Cloud storage for very large data sets and measure their MOM storage in Terabytes across over 50 factories.
Data rates are a bit harder to measure as they vary with time. The rate of reading individual data points from the factory floor will depend heavily on control and business needs. We would not expect to see millisecond raw readings being passed directly to the MOM systems because fast feedback loops are usually needed for direct control and would happen at a level below the MOM system (In, for example, PLCs that are designed for fast access and reaction to data). The MOM system will often require data that business systems and people will react to and timeliness for these would be in the order of a second or slower.
To conclude, MES in the Cloud and on the IoT platform is undoubtedly on its way for many manufacturing facilities. Considering its use should be part of the whole industrial transformation journey.