Prime Minister's Office, South Block, New Delhi
November 1949
---
The CPWD briefing files arrived on Anirban's desk at eight in the morning and occupied the left half of the desk's surface in stacks organized by project category — educational institutions in the first stack, hospitals in the second, the industrial site surveys in the third, and the miscellaneous infrastructure works in the fourth, which was the tallest stack and the one whose contents Anirban had spent the least time with and felt the most guilty about.
The CPWD's Director General, Mehta arrived at eight-thirty with two senior engineers who managed the major project portfolios and a junior official whose function appeared to be ensuring that whenever he needed a specific document, it was already in the room. This was, Anirban had observed over several previous meetings, an extremely efficient approach to senior briefings.
They had two hours. The agenda was progress assessment, budget variance review, and several design questions that had been escalated to his level because they required decisions about priorities that no one below the Prime Minister had the authority to make.
"IIT Kharagpur first," he said, opening the relevant file. "Because it's the good news."
---
The aerial photographs of the Kharagpur site showed something that in another eighteen months would be unrecognizable as the agricultural land it had been in 1947 — a campus that was, in its first construction phase, already beginning to communicate through its physical form the intellectual ambitions that had generated it.
"First phase is approximately eighty-five percent complete," said the senior engineer responsible for the educational portfolio, whose name was Krishnaswami and who had the specific quality of an engineer who had been building important things long enough that his professional satisfaction was genuine rather than performed. "We expect full first-phase completion by the end of February."
Anirban was looking at the photograph. "Walk me through the utility corridor system. I want to understand it the way someone walking through the campus would understand it."
Krishnaswami spread a campus plan across the desk. The plan showed a network of enclosed corridors connecting every major building on the campus in a continuous circuit — a web rather than a grid, with no building more than a three-minute walk from any other building through covered passage.
"The surface level is the pedestrian campus," Krishnaswami said. "Students walk between buildings on the covered pathways at ground level. These are open on the sides — ventilated but protected from rain, shaded from sun. Wide enough for ten people walking abreast. The experience is of walking through a continuous covered environment that connects everything."
He moved his finger to a second layer shown in the cross-section diagram.
"Below the ground level, running beneath every pedestrian corridor and extending to every building, is the utility corridor. This is a separate enclosed passage — sealed, lit, wide enough for a standard laboratory trolley carrying heavy equipment. Every building connects to this sub-level passage through service lifts that open directly into the building's service areas — loading docks, equipment bays, storage rooms."
Anirban looked at the cross-section carefully. "The research equipment transport happens entirely underground."
"Entirely," Krishnaswami confirmed. "A spectrometer arriving from the railway depot comes by truck to the campus goods entrance, transfers to the sub-level through the goods entrance lift, travels the utility corridor to the relevant building's service lift, and arrives in the laboratory without ever entering the pedestrian areas above. A PhD student walking between the library and the laboratory building passes over it without knowing it's there."
He paused for the effect of the design to register.
"The corridor is also the utility spine — electrical, water, communications, sewage, compressed air for laboratories, all the services run through the sub-level corridor in accessible conduits. When a pipe needs repair or a cable needs upgrading, the maintenance team works in the utility corridor without disrupting anything above ground."
"And the future expansion?" Anirban asked.
"Every corridor section was built thirty percent oversized for current load. The utility conduits were specified for twice the current capacity. Every intersection point has provision for a connection spur that isn't built yet but whose foundation space is reserved." Krishnaswami's expression carried the specific satisfaction of an engineer who has designed for the future rather than just the present. "When Phase Two adds the next set of buildings, the utility connections will be made by extending the existing sub-level corridors into the new construction. There's no retrofit, no disruption to the operating campus. The new buildings simply join the network."
"The never-ending corridor," Anirban said. He had described it this way in the original design discussions — the idea that as the campus grew, the covered walkway would simply extend, so that a student arriving in 1980 would walk through the same continuous covered environment as a student arriving in 1950, the campus having grown around them without breaking the fundamental experience.
"We put a lot of thought into the pedestrian network," Krishnaswami said. "The covered corridors are wide enough to be comfortable, proportioned so that even on the busiest days — examination periods, orientation weeks — there's no bottleneck. The intersections are larger than the corridors, creating natural gathering points. We put benches at the major intersections. The experience is of moving through connected spaces that invite stopping rather than forcing through movement."
Anirban was thinking about what it meant for a student to inhabit a campus designed this way — the physical communication of the institution's values in the infrastructure itself. A campus where research equipment moved underground without disrupting students was a campus that had made a specific argument about what was important: that the flow of learning and the flow of commerce were not the same flow, that the student walking between buildings deserved an uninterrupted experience of the campus as a place of thought, that the unglamorous logistics of keeping a research institution running should be invisible to those who were running it.
"The dormitories," he said.
"Single rooms for all students above the first year," Krishnaswami said, moving to the residential block plans. "First-year students are in double rooms, which the design accommodates without making the rooms too small — the room specification is larger than British university standard. Every floor of every dormitory block has a common room, a kitchenette where students can prepare light meals, and a study room with individual carrels."
"The cafeteria."
"Central dining hall capacity for twelve hundred simultaneously. The kitchen is designed for industrial-scale production — we brought in a consultant who had designed institutional kitchens in Switzerland, and the layout reflects that. Multiple production lines so that different dishes can be prepared and held simultaneously. Loading dock at the back for supplies delivery. Cold storage sufficient for a week's inventory. The menu specification—"
He paused, his tone shifting slightly.
"The menu specification was an interesting exercise. We convened a nutritionist from the health ministry, the campus superintendent, and a cook who had worked in a large mess in the navy, and we developed a rotating fourteen-day menu with full nutritional profiling for each meal. Every meal meets specific protein, carbohydrate, and vitamin targets. Regional variation is built in — the menu rotates through the cuisines of different parts of India, so students from Punjab and students from Madras and students from Bengal all encounter familiar food regularly rather than being permanently displaced into unfamiliar eating."
Anirban looked at him. "A nutritionist, a superintendent, and a naval cook."
"We found that the naval cook was the most useful of the three," Krishnaswami said, with the straight-faced precision of an engineer reporting a data point. "He had twenty years of experience producing consistent, high-quality food for large groups in institutional conditions. The nutritionist provided the targets. The superintendent understood the operational constraints. The naval cook understood how to actually meet the targets within the constraints at scale."
"Good combination," Anirban said.
---
AIIMS Delhi had been designed in parallel with IIT Kharagpur, the two projects sharing a design philosophy that had been developed through the same process — specialists brought in to define what excellence looked like in the specific domain, then engineers tasked with building it within the cost and timeline parameters that Indian resources could support.
The hospital was larger in concept and more complex in execution, the complexity of a functioning hospital being a different order of magnitude from the complexity of a university campus.
"Two thousand beds at full capacity," Krishnaswami confirmed. "The layout is what we've been calling the village plan — the hospital as a collection of specialized buildings connected by covered walkways rather than as a single monolithic block. The advantage is infection control — the wards for different specialties are physically separated, which limits cross-contamination risk — and also adaptability. If one building needs to be taken offline for maintenance, it can be isolated without affecting the others."
"The utility corridor principle applies here as well," Sehgal added. "Patient transport above ground on the ward level, all equipment and supplies below through the service corridor. The ambulance bay connects directly to the sub-level so that emergency patients can be moved from ambulance to operating theater without passing through the hospital's general circulation."
"The research institutes," Anirban said. "The Research facilities."
"Adjacent to the main hospital complex but with separate governance space — their own administrative wing, their own seminar rooms, their own visitor facilities. The physical connection to the hospital is through the sub-level utility corridor, which means research staff can move between the institute and the hospital wards without entering the hospital's patient circulation. The boundary is clear in administrative terms and permeable in practical terms."
Anirban ran through the numbers in his mind. Two thousand beds. The largest public hospital in India, by capacity and by design specification. In five years it would be the referral destination for the most complex cases from every AIIMS in the network — the one where the equipment was most sophisticated, where the specialist concentration was highest, where the cases that exceeded every other hospital's capacity would eventually arrive.
"The other four AIIMS," Sehgal said. "Bombay, Calcutta, Madras, Lucknow. All in construction, all tracking to mid-1950 completion for first phase. The design is standardized from Delhi with regional modifications — the Madras facility has greater tropical ventilation provision, the Calcutta facility has elevated ground floor construction accounting for flood risk. Each is twelve hundred beds in first phase with second-phase provision to two thousand."
"And the four IITs in this phase."
"Bombay, Madras, Delhi, Kanpur. Bombay is the furthest advanced — the site in Powai has excellent access and the construction team is experienced. Delhi is the slowest because of the land acquisition complications that were resolved only in March, which compressed the construction timeline. All four tracking to mid-1950 for first-phase completion."
---
The Gurukul Sanghatan presentation was longer and required more of Anirban's attention because the decisions about the flagship school's specifications had been delayed pending his direct involvement, and the decisions about the district school program had generated questions about standardization versus adaptation that required policy direction rather than just engineering judgment.
Krishnaswami spread the flagship school's master plan across the desk.
"Fifteen hectares for the campus. Roughly fifty thousand square meters of built area for the first phase, with expansion provision to sixty-five thousand."
He began working through the components in the methodical sequence of someone who had been building this in his mind for months.
"The academic complex is the heart — classrooms, laboratories, staff offices, grouped around a central quadrangle with covered arcades. The science complex is separate, positioned downwind of the dormitories because of ventilation requirements, with dedicated laboratory space for physics, chemistry, and biology, each with full bench length for thirty students working simultaneously."
"The engineering workshop complex," Anirban said. "Size?"
"Larger than most schools would build. We're treating it as a genuine fabrication facility — machine tools capable of actual production work, not demonstration equipment. A metalworking section, a woodworking section, a foundry capable of small-scale casting, and a drafting room. The rationale is that students at Gurukul schools should be able to produce real things, not just understand how things are produced."
Anirban nodded. This was the principle he had insisted on — that the Gurukul schools not merely teach science and engineering as abstract knowledge but provide the physical context in which that knowledge was applied, so that a student who graduated understood not just the theory of mechanisms but the experience of building one.
"Library," Krishnaswami continued. "We specified for fifty thousand volumes, which is larger than most public libraries in India currently. The stacking arrangement allows for expansion to a hundred thousand without structural modification. Reading room capacity for three hundred — carrel seating, not table seating, because individual study work requires acoustic separation."
"The dormitories."
"Eight blocks for students, two blocks for staff. Student blocks at one hundred and fifty per block, giving twelve hundred student capacity. Each block has house parent accommodation — not separate housing but a dedicated flat within the dormitory block, which places the house parent in the building at all times. Every floor of every block has a common room and a study area. The dining complex serves the whole school — a single large hall rather than distributed messes, because communal eating for the full school is part of the social architecture."
Anirban looked at the athletics facilities on the plan. "Four hundred meter track."
"Synthetic surface — we brought the specification from the SAI . Adjacent to the track, the swimming facility: fifty-meter pool, eight lanes, covered but naturally ventilated. The football and hockey pitches share a ground — the hockey surface requires specific drainage that takes priority, and football adjusts to the same surface. Basketball and volleyball courts in the gymnasium courtyard, which is covered for all-weather use. The gymnasium itself is equipped for strength training and gymnastics."
"Agricultural plots."
"Three hectares designated. This was a specific requirement from the education ministry — every Gurukul campus maintains agricultural land that students rotate through in their curriculum. Practical agronomy, soil science, irrigation management. The plots are genuinely productive — the harvest supplies a portion of the dining complex's vegetable requirement."
Anirban thought about the scale of what was being described. Three thousand students on a fifteen-hectare campus with fifty thousand square meters of built area — and this was the flagship. The district schools, the standard model that would be built in every district of India, were smaller by specification but not by ambition.
"The district school standard," he said.
Krishnaswami produced a second plan, less detailed than the flagship but following the same organizational logic at reduced scale.
"Eight hundred to twelve hundred students, depending on district population. Approximately twenty-five thousand square meters of built area. Four dormitory blocks rather than eight. The athletics facilities are complete but not at the same scale as the flagship — three hundred meter track, standard pool, combined sports ground. The workshop complex is proportionally smaller. The library is specified for twenty thousand volumes."
He looked at Anirban directly.
"One per district of India. We've been doing the arithmetic. India has approximately two hundred and fifty districts at present, with the integration of the extended territories bringing the number to approximately three hundred. Three hundred Gurukul schools at the district standard is a construction program of significant duration."
"Twenty-five years," Anirban said. He had done the arithmetic himself, many times. Three hundred schools was not a ten-year project; it was the work of a generation. "The first cohort — the districts where the social and economic rationale for priority is strongest — is being identified. We begin there and we build outward. The construction program runs continuously."
"That has implications for the construction industry," Mehta observed, with the expression of someone who has been building to a point. "A continuous program of Gurukul construction at the district level, running alongside the IIT and AIIMS program for the next decade, alongside the Kendriya Vidyalaya program, alongside the university campus construction — this is a sustained construction demand at a scale that doesn't exist in current Indian construction capacity."
"Which is why we are building the capacity," Anirban said.
---
The Kendriya Vidyalaya program presented a different picture — lighter specification, faster construction, and a standardization that allowed what the Gurukul program could not: the production of a design once and its replication many times with minor local variations.
"The KV standard school," Krishnaswami said, spreading the much simpler plan, "is approximately eight thousand square meters. Classrooms for twelve hundred students, a library of five thousand volumes, basic science laboratories, a playing field for standard sports, modest staff accommodation. The construction time for each school, using the standardized components we've developed, is seven months from foundation pour to completion."
"Sports facilities."
"Standard — football and cricket field shared, a covered area for indoor games, basic fitness equipment. We deliberately did not specify the swimming pool or the gymnasium or the full athletics track that the Gurukul specification includes. The rationale—"
"Is that the Gurukul schools serve students who will be identified for them through merit selection and who represent the program's investment in future leadership," Anirban said. "The Kendriya Vidyalayas serve the children of government servants who need continuity of schooling quality, not specialized development. Different mandate, appropriate specification."
"Yes," Krishnaswami said. "Which is why construction is progressing at a rate that the Gurukul program, with its complexity, cannot match. We will have two hundred and ten Kendriya Vidyalayas operational by mid-1950. The target for 1952 is five hundred."
---
"Now the other part," Anirban said, pulling the industrial and PSU reports into his line of attention. "SAIL. Walk me through the Ranchi backward integration logic."
Mehta produced a report that was denser with numbers than any of the construction files, the financial modeling of the Steel Authority of India's current operating strategy presented in tables that required a specific kind of reading.
"The main plant is not yet producing steel from raw materials," Sehgal said. "That is on track for late 1950. In the interim, SAIL identified a way to generate operating revenue rather than simply consuming capital."
He explained it as Anirban had understood it from earlier briefings but which he wanted confirmed: the semi-finished steel market in India was substantial, fed by imported ingots and billets that arrived from British and Belgian and American mills and were then processed into finished products by smaller fabricating operations. SAIL had positioned itself to buy these semi-finished imports at wholesale prices and process them through the workshop facilities that were already operational on the Ranchi site — the facilities that had been built first precisely because they were needed for the plant's own construction — converting billets into structural sections, ingots into plates, and selling the finished products to the construction boom that SAIL's own existence was partly driving.
"The margin," Anirban said.
"Between imported semi-finished and finished structural section prices, approximately twelve to eighteen percent. SAIL's processing cost is below this margin because the labor cost in India is significantly below the European mills that sell the finished product at imported prices. The result is that SAIL is currently generating operating surplus rather than consuming working capital while the main plant is under construction."
The elegance of this was the kind of thing that made Anirban appreciate competent industrial management — the ability to identify that the construction of the thing and the operation of the thing were not sequentially separate but could be overlapping, that the workshop capacity built for the plant's construction could be turned to commercial operation in the interval before the full plant was operational.
"And the surplus funds the remaining construction," Anirban said.
"Part of it. Combined with the capital budget, the construction is now effectively self-financing in the sense that the capital requirement from the NIIF has been reduced by the operating surplus. The finance ministry's revised projection is that SAIL reaches the main plant commissioning stage with its capital allocation approximately eighty percent utilized, having funded the remainder through operations."
"The Cement Corporation."
"Steady," Sehgal said simply. "The construction boom is a cement market that exceeds the Cement Corporation's current capacity by twenty percent. They are rationing supply to priority projects, which means the priority designation for educational and health institutions actually operates as an allocation mechanism in a constrained supply environment. The profit margin is good precisely because demand exceeds supply."
"They need to expand."
"They are expanding. A second plant in Rajasthan is under construction, funded from operating surplus. The expansion is self-funding from the same logic as SAIL."
Anirban looked at the pattern that was forming — the PSUs generating their own capital for expansion from the margins created by a construction demand that was itself being generated by the government's development program. The NIIF's initial capital was not being consumed by ongoing operations; it was seeding the capacity that generated operating surpluses that funded further capacity that met further demand that the development program was creating.
The flywheel was turning.
ONGC was profitable in the early stages, supplying refined products to the construction fleet and to the industrial operations that were running on diesel and industrial lubricants. The other public sector utilities — the National Housing Bank providing construction finance, the industrial banks channeling credit to the private contractors who were building the schools and hospitals alongside the PSU construction firms — were operating in the black rather than drawing on government support.
"RVNL and NBCC," Anirban said.
"The rail construction and urban building companies are running at full capacity and have been unable to expand fast enough to meet the demand directed to them," Sehgal said. "This is actually where we have the most significant constraint in the construction program — not material supply, which is adequate, and not design capacity, which the architecture and engineering firms have scaled to meet demand, but pure construction management and skilled labor at the supervisory level."
"The private players."
"Taking the overflow. L&T, Gammon, Tata Construction, several other firms that have expanded their workforce and their equipment fleet in response to the sustained demand. Their margins are good because the market is a seller's market — there is more construction work than construction capacity, which means the firms that can reliably deliver are able to command rates that allow expansion rather than merely survival."
"Are they expanding fast enough?"
"Not quite," Sehgal said honestly. "The skilled labor constraint — the site managers, the structural engineers, the quality assurance supervisors — takes time to resolve because you cannot create experienced construction management quickly. We are funding training programs through the CRC technical posting scheme, but the people completing those programs will be ready for junior roles, not senior site management. The senior people are being developed through the work itself, which is the natural mechanism, but it is the slow mechanism."
"Three years to solve," Anirban said.
"Approximately."
"Then we accept the constraint for three years and ensure that the priority allocation of construction capacity consistently favors the highest-impact projects." He made a note. "Any projects that are slipping due to construction management constraints should be flagged specifically, so that the allocation of the available capacity is directed to them rather than allowed to drift."
There followed a period of approximately forty minutes in which the discussion moved through a range of topics that were simultaneously important and fundamentally unengaging to anyone who was not directly responsible for them — procurement specifications for hospital equipment that had been approved but not yet contracted, the land acquisition status for six Gurukul school sites where the compensation disputes were in various stages of administrative resolution, the design modifications required for the Calcutta AIIMS ground-floor elevation, the scheduling of the construction inspection visits for the completed first-phase IIT buildings, and the documentation requirements for the Ministry of Finance's quarterly capital expenditure reporting.
Anirban worked through all of it with the same attention he gave to the things that he found genuinely interesting, which was the discipline that distinguished governance from performance. The procurement specifications mattered because the hospital equipment that arrived would either work or not work and the difference was determined in the specification. The land acquisition status mattered because a school site under dispute was a school that wasn't being built. The ground-floor elevation mattered because a hospital that flooded was worse than no hospital.
He signed the relevant approvals, directed the questions he was not empowered to answer to the appropriate ministry channels, and made notes about the follow-up items he wanted tracked.
By eleven o'clock, the meeting was concluding and Sehgal's team was gathering their materials with the efficiency of people who had accomplished what they came to accomplish and were returning to the work that the meeting had generated.
"The IIT Kharagpur opening," Anirban said, as they were leaving. "When?"
"First of January, 1950," Krishnaswami said. "First cohort of students — five hundred undergraduates and two hundred postgraduates. We'll be ready."
"Good," Anirban said.
After they left, he sat with the campus plan still spread across the desk — the web of corridors, the utility sub-level shown in dotted lines beneath every pedestrian pathway, the dormitory blocks in their configuration around the dining complex, the athletics track with its synthetic surface, the engineering workshop marked with the specific annotation that distinguished it from a teaching laboratory: production-grade equipment, not demonstration grade.
A student who walked through that campus for four years would emerge knowing that excellence in engineering was not only about understanding forces and materials and mechanisms but about the specific physical experience of making things — the feedback of the lathe, the smell of machining oil, the noise and the heat and the precision that the workshop required.
He thought about the students who would inhabit the dormitories whose common rooms had been designed to invite staying rather than passing through, who would eat in the dining hall whose menu had been designed by a nutritionist and a superintendent and a naval cook, who would walk through covered corridors over the utility passageways where research equipment moved invisibly beneath their feet.
They would be the engineers and scientists of the 1960s and 1970s. They would build the things that would make India in 1980 unrecognizable to anyone looking at India in 1950.
They did not know this yet.
But the buildings that would form them were already taking shape in Kharagpur, in Powai, in Adyar, in Hauz Khas, in Kanpur. The physical environments that would shape what they could imagine and what they could attempt were being poured in concrete and laid in brick, connected by covered corridors and utility passageways, specified to the standards that a naval cook and a Swiss kitchen consultant and a nutritionist had together determined were appropriate for young people who were being asked to carry India's scientific future.
The flywheel was turning.
The building season was in full execution.
And somewhere in Kharagpur, the last ten percent of the first phase was being completed in the November light.
By January, there would be students.
The algorithm would have new inputs.
And the output would take twenty years to assess.
That was the nature of the enterprise.
It required faith in arithmetic.
Which was, in the end, exactly what engineers were trained to provide.