Story: A Hook-Rug Company

Story: A Hook-Rug Company

Aniruddha Sarkar

06 May 2020

(Man, behind the machine: the story of a private data network design for a PSU Iron & Steel major while learning a few lessons on traffic analysis and capacity sizing, both so fundamental to any solution design)

On an international flight, one gentleman asked his co-passenger, ‘Oh, you are from India. Can you supply me a few Hook Rugs?’ The co-passenger, one of our organization’s representatives, said ‘Yes’. They did some quick negotiations and agreed upon a price of USD 2000 apiece. The gentleman then asked, ‘Do you have questions?’ Our rep said, ‘Yes Sir. What is a Hook Rug?’

Here is a sublime message: ‘Always say yes to a customer, then find out his requirements’.

Our organization would do anything that a customer wanted. What the large corporations in India were looking for during the mid-1980s?

That was connectivity.

Our organization rolled out a public data network solution, INDONET. The name says it all. Within a few days, our organization started receiving calls from across the industry segments. All big corporations with operations scattered in many parts of the country needed connectivity of their own, but these corporations were clueless about how to go ahead.

With a public data network already up and running, our organization next formed a private network group in the mid-1980s. A galaxy of experienced professionals from various industries, at home and abroad, joined this group. Also, some hard-core expert professionals from leading Indian research organizations, telecom companies, and defence establishments joined this group. The organization started providing consulting services to all who came calling. That was an exciting time.

While forming various consulting teams to address industry-specific private voice cum data network requirements, our organization realized that it already had many electronics and telecommunications engineers who could supplement this private network group.

One day, our Eastern Region Systems Integration(SI) Division Head pulled me out of a large port automation project. This project was in its design and development phase.

A team of experts from our New Delhi-based Corporate Private Network Group was visiting Kolkata. I, along with three more colleagues from the Eastern Region, joined this team.

We travelled by an early morning train to Dhanbad on our way to a large integrated Steel Plant in the neighbouring State of Jharkhand.

We had one expert from the US, a Ph.D. from the California Institute of Technology, and a satellite communications technology expert. This gentleman was an industry veteran from Bell Lab, USA. He had set up a small consulting company of his own in New Delhi. Our organization had hired him as an external consultant.

Our organization initially planned this project under the corporate project account. The corporate marketing team sold this project at the PSU headquarters at Lodhi Road, New Delhi.

But Eastern Region Systems Integration(SI) Head argued that all the major steel plants of this PSU giant company and their captive mines belonged to Eastern Region (ER) geography. So, the ER team should execute the project.

Then the Corporate Team raised a crucial point. They asked whether the region could engage the right Telecom Experts from the region. Eastern Region said, yes. Thus, ER management included four professionals in the team, all from ER. We all were telecom engineers. One very senior colleague had ten+ years of experience in the Customer Services (CS) division. I already gathered six years of experience in Systems Integration (SI). Two other colleagues were fresh graduates from the Indian Institute of Technology. One of them was also an MBA from IIM Ahmedabad. Both of them belonged to the ER SI division.

With such profiles of staff members from ER joining the project team, there was no further issue. Eastern Region became the project owner.

Our ER SI Division Head briefed us about this tug-of-war and said, I fought for this project and got it for our region. Please put in your best efforts so that we can establish our competence within the region to handle such assignments. He also mentioned that there were many such private network design and implementation projects in the pipeline.

We reached the large Integrated Steel Plant in today’s Jharkhand by mid-day and presented the concept of a wide-area network that could carry both voice and data traffic. The idea of connecting all steel plants, mines, major city-based regional marketing offices, divisional offices, branch sales offices, stockyards, etc., was a revolutionary concept for this PSU audience then, in December 1986.

We proposed a dedicated high-speed backbone satellite communication data-cum-voice-network. This wide area network (WAN) would be based on state-of-the-art technologies. It would connect all the computers, all voice communication devices, and all relevant end equipment across the customer organization within the country.

Since the Captive mines required local connectivity with the nearest railway sidings, we proposed ultra-high frequency (UHF) radio communications.

This integrated voice cum data network concept was a paradigm shift for the audience who were senior officials of the steel plant. They showed a keen interest in this initiative.

After deliberations in the same forum, we proposed the following approach:

A team from our organization would circulate a questionnaire in advance. Then we would conduct a follow-up voice & data traffic survey visit to a few selected all-India locations of the PSU. We would gather information on voice and data traffic volume. We would then process that information to arrive at the network capacity sizing data. Next, we would design an appropriate voice cum data network.

The Customer agreed. Subsequently, we and the customer jointly prepared a plan of action for an all-India site survey.

With this presentation, we achieved a grand purpose. We set the road map for the project activities. We could also sensitize the key stakeholders of a steel plant about the project and get buy-ins. After another round of presentations to the Central Marketing Organization (CMO) HQ of this PSU customer at Kolkata, our Delhi-based team left.

Next week, we planned, finalized, and distributed the questionnaire format. The target number of locations worked out to be around Three Hundred and Sixty.

We, the four members of the ER team, worked out a strategy for the voice and data traffic volume data gathering at customer locations through physical visits. We also shortlisted a few locations based on some criteria in consultation with the customer.

Two younger colleagues planned for a trip all over south India. We, the two senior colleagues, planned to cover north India.

Our approach was like this:

We would first establish contact with all the representative customer locations through proper channels and then send a copy of the questionnaire to facilitate their prior readiness with the data we would look for.

Next, we would plan a visit, and communicate in advance the date and time of our visit to a location. Then we would physically visit the location and conduct a day-long study.

Within the next three months, we crisscrossed the country. First, we visited many representative locations of various categories having major and minor communication traffic volumes. We collected data through individual interactions. Next, we merged that data. We also received a few feedback by mail.

(Just a personal note here. I later computed how many kilometres I had to travel for the initial three months of the project. It summed up to a staggering thirty-seven thousand km!)

Next, we conducted the analysis and extrapolation of data to all the three hundred and sixty-plus locations. We computed the projected daily, and hourly peak voice calls data for the next five years in Erlang (a unit of measurement for voice traffic load).

We also conducted a similar analysis for the data traffic projection.

In the end, we performed sanity checking along with the key customers to make sure that data quality was reasonably good.

We now prepared ourselves for the capacity sizing of all the communication links. The satellite transponder bandwidth requirements had to be computed.

We also formally submitted our traffic analysis reports in multiple bound volumes to the consultant.

Here, one interesting tussle happened between our organization and the consultant.

One day, the consultant handed over a draft detailed project report (DPR) to us. We, the ER Team, were then camping in New Delhi. We had a feeling that the consultant was not seriously considering our traffic analysis data projections. His assumptions were much less than our estimates.

Initially, the consultant was downplaying our findings. He said there would be an enormous cost for that kind of transponder capacity.

Our private network group seniors stuck to our traffic analysis figures. After all, we did a massive data gathering, data sanity checking, data analysis, forecasting, etc., using mathematical models.

Finally, the consultant had to rework the whole design over again. It took some more time. Throughout this period, we, the ER team along with our corporate team, would visit the premises of the consultant to provide any clarifications on the data or clarify doubts.

The consultant had his model design software, where a large volume of traffic data was required to be populated against every single path of communication. The maximum communication paths were a mind-boggling 365 X 365 = 1,33,225! As per findings, around 40% of these paths were carrying minor to major traffic. We could understand why the consultant skipped this non-trivial exercise initially.

On our insistence, the consultant finally populated all the path-wise traffic records over a period. The actual picture emerged. Now, the model software could generate a design that looked reasonable. It could also generate a Bill of Material (BOM).

Our corporate team of experts did individual checking based on their area of expertise in various related knowledge areas. The consultant also resolved a few technology-related issues raised by our corporate team. That again affected the BOM.

After several passes of checking and verification, rework, etc., the project team could complete the BOM. We could thus arrive at the final project implementation cost.

We, the team of four engineers from the ER, really felt happy about this last episode. The revised DPR was now ready based on the traffic projection data suggested by us!

Our organization submitted the final DPR to the customer along with all annexures comprising our analysis of traffic volume supported by survey datasheets signed by the customer’s ‘authorized signatories’ from all India locations.

The customer, in due course of time, implemented the solution using NICNET backbone infrastructure, which was by then just operational. Our organization, being a consultant, did not take part as a solution implementer.

Learning:

Traffic data gathering, analysis, and corresponding sizing of a solution can be a discipline by itself. It may apply to many knowledge areas or problem domains. Traffic may be (1) voice traffic, (2) network data traffic, (3) computer processor task load traffic, (4) database table-level, record-level, data Item-level access read or write request traffic, (5) a buy or sell request traffic on a scrip at peak trading hour in a securities exchange that influences the trading engine processor capacity sizing, and many more.

Please find below a few professional hazards faced during the nationwide survey:

On our return from a Branch Sales Office visit in Mandi Gobindgarh, Punjab, in early 1987, we faced a few locals on our way back who quizzed us for some time and then let us return to Chandigarh. During those days, Punjab witnessed some disturbances. The local car driver later told us, after returning to Chandigarh, that we had a providential escape from death.

On a visit from Agra to Gwalior Branch Sales Office, we had to stop our car at a barren place in Chambal and ask for directions. The group of people sitting on a large boulder had guns, which I noticed later. They asked about our place of stay, and the purpose of the visit, and then showed us the right direction. Later, on return to Agra, the driver opened up. Those people were the dreaded bandits from Chambal.

After a very hectic trip to a steel plant and mines for three days, we booked first-class tickets for an overnight passenger train journey from Durg to Nagpur. The next early morning we had an appointment at the Divisional Office in Nagpur. At the Durg Railway Station, close to midnight, somebody locked the first-class compartment from inside. We tried in vain to get inside. Finally, we got into a luggage van carrying parcels and had to spend a sleepless night's journey for ten long hours.

In Dimapur, Nagaland, we could not locate the Branch Sales Office at the right address. Later, we could notice a small handwritten direction on a staircase. We climbed up to the first floor. The official concerned was terrified because of some security-related incidents the day before. We got all the information as per the questionnaire. Later, we also followed a circuitous route to reach the airport. The main road remained blocked for some local issues.

Sharing a few pleasant memories:

On one early morning, we both colleagues started from the steel city of Rourkela to an iron ore mine some 100 km away. After a long three-hour drive through hilly terrain and jungle, we reached the mines. We completed the survey in another four hours. Then we started our return journey well past 2 pm. Next hour, there was no human habitation in sight. We were feeling hungry and exhausted. Suddenly, we traced a tribal home and stopped there. Within an hour, the old man and his wife served us hot cooked food with local vegetables and delicacies, like fresh fish curry. The taste of that food and their hospitality are still fresh in my mind.

We travelled during a survey by an early morning flight From Delhi to Patna via Lucknow. We witnessed the majestic peaks of Garhwal and Kumaon Himalayan Ranges all along. After field study at Patna, we took the next morning flight from Patna to Guwahati via Bagdogra. Again, we got beautiful glimpses of Everest and Kanchenjunga Peaks of the Himalayan Ranges. During the Guwahati to Dimapur flight, the views were equally fascinating. These were experiences of a lifetime.

-----x-----