There will be selective adoption of automation in some areas
Published on : Saturday 22-02-2020
Many companies of repute have invested in building digital roadmaps.
Raghavendra Rao, Associate Partner & Senior VP, Frost & Sullivan, in conversation with Milton D’Silva.
One of the biggest challenges in implementation is procrastination, or waiting for the right moment. How best this can be addressed?
When it comes to investing in cutting edge technologies, which is capital intensive, there is a niche of MNCs and larger companies with deep pockets who opt for the latest – IIoT, Artificial Intelligence, you name it. Then there is the next level, the large companies which would start by preparing a digital roadmap; but ultimately how much they invest and in which areas, depends on a couple of things like the immediate need and the RoI. The comparison is on cost – whether to invest in automation or continue with manual operations as that is cheaper – and to that extent, we may not have the level of automation adopted by developed countries. Also schemes like NEEM (National Employment Enhancement Mission) for practical job training address the issue of skill development. But there will be selective adoption of automation in some areas like customer centric activities or even personal safety, with deployment of smart robots or the cobots.
Can this then become an incremental thing? What stage are we in this journey?
There are stages in maturity – Stage 0, Stage 1…Stage 5. At Stage 0 the machine is purely mechanical. At Stage 2, the machine has CNC but is still not connected, just standalone. At Stage 5, the machine is truly networked at the organisational level and all data is fed to a central server, is processed in an intelligent way, sent for predictive analysis, and the feedback used to further improve machine performance. So in general, we are not there yet, with perhaps some notable exceptions. However many companies of repute have invested in building IIoT or digital roadmaps, basically for management to understand what it implies. There is a fair bit of intent on the management’s part to do this, so they have engaged consultants. I suspect most companies are at this stage of evolving. How deep they would go, when and where they start investing –that’s where we stand today.
While we are not even there in terms of full automation, there are a couple of recent examples, notably the Adidas speedfactory, going back to regular production?
Well, this is not something that can be generalised based on one particular instance. There could be such cases, within the larger scheme of mature things, areas which should not be touched by automation as that is better served with human element in it. It is like you decide to go for full automation and then while in the process, realise there are some things that require human touch, which too is part of the evolution. I would not find fault with that or say it is failure of automation, it is possibly a learning process. Where there is low availability of manpower, full automation is the need. Take the example of the car, which is evolving continuously and is now preparing for autonomous driving, eliminating the need for a driver. But there would still be cars that are driven manually. So even with full automation, there will still be need for skilled manpower. The requirement could be lesser, as one person can oversee a greater number of consoles or stations, e.g., refineries, which today are among the most automated places.
Talking of refineries, we have the example of Reliance Jamnagar. But even the public sector refineries now are adopting automation at greater levels.
In refineries the core focus is on safety parameters, so automation is the way to go. There are also the high highs and low lows, different crude grades that need different processes, the checks and balances, the control factors. Investment in technology would ultimately mean how effectively you can use low cost changes to effect highly profitable outcomes. Factors like how much more yield can be obtained, the sulphur content, the quality and above all, the right product mix according to the market needs, producing high margin products on demand. The top refineries are now investing in Stage 6, which is the next stage in maturity. All refineries today are automated, but are at different stages. The Reliance Jamnagar is perhaps the most automated where no one has to step outside the control room. But others may have some functions requiring some manual intervention – say open some valve, read the flow once before automation takes over.
We have a situation today where to compete globally we need more automation, but local factors affect the deployment. Is there a way out of this?
The key is in identifying which parts of the process need to be automated. There are rules followed in automation for discrete manufacturing, and there are various schools of thought. Like processes that are below 5 seconds, or a customer quality sensitive process, and of course a safety hazard for workers, in these three cases you mistake proof or Poka-Yoke the process so that an error does not occur; or if it occurs, does not go unnoticed. So there is an evolution of certain algorithms which will define the process to automate first – and Indians are experts in this– they will say, look, this system costs USD 5 million, but I have just USD 1 million, so please tell me as the expert, where should I put this money in my process, as I do not have the resources to automate everything. This way, he gets the benefits of using automation, but at a lower cost, and still manage the efficiencies relying on the manual skills to operate those areas that are efficiently managed the old way. That’s how it happens; I do not see people splurging money just because technology is available. The entrepreneurs here are good managers of people and resources.
Let’s talk about predictive maintenance, where IIoT offers low cost options, but these may not be highly reliable?
This is a conscious choice a company has to make. Without sounding generic, I would again refer to refineries. Reliance Hazira – also Jamnagar, but the former is more complicated – is an example of very good predictive maintenance. But this does not come cheap, there is a cost. Today if you ask me, most Indian companies are happy with the TPM way of working, which means operators doing all the maintenance, following the daily and weekly checklists, which is not predictive but more of preventive maintenance. But again, this depends on the criticality factor as critical machines are the ones that need predictive maintenance, like say the kiln in cement industries. Predictive maintenance is based on reading parameters like noise, heat, vibration, etc., which is done by sensors. Now any company with production constraints will not ignore this and invest in predictive maintenance, but not all do. There are those with enough installed capacity for whom a machine going idle is not a big deal. That is the reality.
