Alum Navigates Systems Challenges to Launch Successful Milk-Chilling Business

Sorin Grama, SDM ’06

By Sorin Grama, SDM ’06

The challenge: Milk is India’s lifeblood. Indians depend on milk for much of their daily nutrition. It is used in curries, the beloved chai, and even for religious rituals. India draws its milk supply from millions of small farmers in villages scattered across the vast countryside. Milk must be collected twice every day, 365 days a year, and rushed to a processing center before it spoils. As a result, the milk supply chain presents a huge challenge for dairy processors.

I learned about this challenge in 2007 when I was visiting India for the first time looking for business opportunities. One of our hosts was a dairy in Bangalore that was having a problem collecting fresh, quality milk. I learned that milk is collected in three steps:

  • Step 1: Individual farmers deliver 5–10 liters of milk to a collection center in a village. A collection center may aggregate from 500 to 2,000 liters of milk per day from 20–40 farmers.
  • Step 2: Milk is picked up and transported to a nearby chilling center. Because refrigeration is not used at the village collection center, the dairy processor needs to pick up the warm raw milk quickly, within 5–6 hours, before it spoils.
  • Step 3: The milk is transported from the chilling center to a processing center where it is pasteurized and processed into such products as cheese and ice cream.

These steps are repeated for thousands of collection centers, twice every day, all over India. It is a huge logistics challenge, and one that I soon discovered could have significant business potential.

One of the farmers who uses Promethean Power Systems’ equipment carries a jug of milk in the village of Mottur in Tamil Nadu, India. (Images courtesy Promethean Power Systems.)

The approach: I began by studying the collection process fully so I could identify the pain points. In 2008 my team and I spent an entire month traveling through rural India following the “milk trail” from farmers to consumers. We made a video of this process so we could later explain the challenge to our US partners and investors.

We learned that the highest pain point in this supply chain was at the source, in the villages where the milk is produced. If milk is not refrigerated immediately after milking, bacteria starts to grow exponentially, changing the taste and eventually spoiling the milk. The sooner milk could be refrigerated, the better it would be for everyone in the system: farmers, processors, and consumers. If milk could be refrigerated at the village, multiple benefits would accrue, including:

  • Lower transportation costs. Milk could be picked up just once a day.
  • Access to additional supplies. Since refrigerated milk lasts considerably longer, the supply chain could extend farther into the countryside.
  • Better milk quality. This benefit is particularly significant since it would allow processors to sell higher-value milk products such as butter, yogurt, and ice cream.

    Figure 1. Traditionally, milk collection in India follows three steps (top image), which means fresh milk typically goes unrefrigerated for several hours—during which time bacteria can grow and spoil the milk. Promethean Power Systems was launched to facilitate a two-step process (bottom image) that refrigerates milk much sooner, curbing the growth of bacteria and preserving milk quality.

    If there are so many benefits to refrigerating milk at the source, I had to ask: Why weren’t dairies doing this? The answer is simple. Refrigeration in rural India is difficult to achieve because of an underlying problem: lack of reliable grid power. The milk supply challenge is really a power infrastructure challenge. If a refrigeration system could be reliably powered, the main problems in this supply chain could be addressed.

    The tools: To understand the problem more deeply, I used contextual inquiries and immersion in my customers’ world, two user-centric design methods I learned in Product Design & Development, a course I took while a student in MIT System Design & Management (SDM). Based on these observations, I decided the best solution would be a stationary milk chiller that could be operated at a village collection center to chill milk immediately after it has been delivered by farmers.

    To design this solution, I used the systems architecture and product design teachings that were still fresh in my mind at the time. I tackled the problem by first decomposing the system into modules that could be designed and developed separately. I then integrated these modules into a final system.

    The two critical modules in this system are: the power module and the refrigeration module. For the power module, I quickly determined that solar power would be the best option. I had some experience with solar power—my SDM thesis was a survey of thin-film solar technologies—and believed that solar power costs would come down dramatically over time. Next, I began to explore different ways to achieve refrigeration efficiently using solar power. I investigated thermoelectrics, absorption chillers, and iceboxes, but I eventually settled on DC-powered vapor compressors.

    Figure 2. Fixing on solar power made it possible to explore multiple concepts for refrigeration.

    The results: My team and I built three prototypes over a period of three years, and with each prototype we learned something new and improved the design. However, the cost of solar power was still prohibitively high, a challenge exacerbated by the difficulty of obtaining the DC components we needed at a reasonable price. The most painful and vivid lesson came when our pilot customer rejected the system as too complex, too expensive, and too difficult to install. We had spent all our time and money to design a beautiful solar-powered refrigeration system only to discover that it was impractical and uneconomical. It was a tough lesson, and it looked as if it would be the end of the road for our startup.

    But there was a glimmer of hope…

    During the process of designing the solar power module, I had also designed a backup subsystem, since solar is not very useful unless you can store the energy and use it later. Because we were dealing only with refrigeration, I chose a simple thermal storage system comprised of a cold water tank. During the day, water can be chilled and stored in an insulated tank. The chilled water can then be used to chill milk in the early morning and late evening. The thermal backup was almost an afterthought, a necessary but not a critical component.

    With a bit of reflection, I realized that we could drop the solar component and instead use the existing power grid with our thermal storage as a backup. The grid is usually available in villages; it just doesn’t always work when you need it. We could charge our thermal battery when the grid was on, and use it as a backup when the grid was off. It was a simpler and more elegant solution. We quickly built a prototype, tested it, and it worked.

    From then on things moved quickly. We iterated and improved on the thermal storage system, which became our differentiator and the source of our competitive advantage over conventional milk chillers, which use diesel backup generators—an expensive option. We eventually patented it and used it for other cooling applications. To date, we have installed more than 600 chilling systems throughout rural India. Each system has a capacity of 1,000 liters and serves the needs of 30 to 40 farmers.

    The lesson: With the benefit of hindsight, I realize the mistake I made during my system design. To be fair, it was a complex system with a lot of moving pieces. I started by fixing on a power source (solar) and investigated different concepts for refrigeration. What I should have done is fixed on a known and economical method of refrigeration (AC-powered vapor compressor) and explored different concepts for generating reliable power: solar, biogas, battery, etc. After all, this was a power infrastructure problem that I was solving, not a refrigeration problem. My bias and preference for solar prevented me from truly exploring the full solution space for this problem. I learned my lesson the hard way, but I don’t regret the journey. Mistakes were costly, but they were also sources of inspiration.

    Figure 3. Ultimately, it became clear that the solution space for India’s milk-chilling challenge should have included additional power source options.

    About the Author

    Sorin Grama, SDM ’07, is the cofounder of Promethean Power Systems, which manufactures and sells milk-chilling systems in India, Bangladesh, and Sri Lanka. After living in India for a few years, Grama is now back at MIT as entrepreneur-in-residence at the Martin Trust Center for MIT Entrepreneurship and the Legatum Center for Development & Entrepreneurship.

    For more information about Sorin Grama, SDM ’07, and his company, Promethean Power Systems, visit