IoT Builder Spotlight: K Sridharan
Zan Compute: AI for Smarter, Cleaner Facilities
IoT is already beginning to fulfill its potential to transform human life for the better.
Connected devices make it possible to protect crops, monitor equipment, ensure supply chain integrity, and secure food safety — to name just a few use cases — at a pace, level, and cost threshold that would not otherwise be possible.
Globally, there are very few industries that have established such strong alignment between profit motives, social outcomes, and human health objectives from the 17 United Nations Sustainable Development Goals. Sensors make our world stronger, more resilient, and smarter. Meanwhile, the global IoT market is growing thanks to advancements in ancillary tech such as AI.
According to the World Economic Forum, “In the new, unpredictable world, the internet of things gives us the real-time awareness we need to adapt and thrive.”
So what does it take to start a company — and build technology — that directly impacts human health and wellbeing? To explore this extremely important question, we sat down with K (Sri) Sridharan, chief product officer at Zan Compute, a company that provides facility management services to commercial buildings.
After 26 years at Intel, Sridharan realized the time was right to commit to building IoT solutions as an entrepreneur, and joined Zan Compute started by a long-time Intel colleague. With a passion for innovation and a lifelong dedication to improving human health, he helped develop a company that dramatically reduces inefficiencies in keeping buildings clean. In this interview, we learn about his path to entrepreneurship, thoughts about product development, and perspective into the evolution of IoT.
His underlying mission is to create something of value, that solves real problems and improves lives in a big and practical way. He has aligned this altruistic vision with a clear, defined, and expansive market opportunity.
What was your path to entrepreneurship like?
Innovation has always been around me here in Silicon Valley. At Intel, I developed a passion for the application of technology rather than the technology itself. It was an incredible environment, and I was lucky enough to join some interesting projects across multiple product groups. We had the opportunity to build ideas from scratch and then take these innovations to market. It’s this environment that kept me at Intel for so long.
It was here that I met my colleague Junaith who worked with me on context aware computing, personalization, and recommendation systems — as well as sensors. He addressed these problems from a research perspective, and I supported his group in taking this research to market. My group’s focus was around pragmatic questions — how to translate an idea into a product, identify a market for it, and figure out pathways to iterate on ideas.
Towards the end of my journey, corporate life started becoming mundane — things were getting a bit same old, same old. Junaith left Intel in 2014 to start Zan Compute, and he invited me to become an advisor to the company. Two years later, the perfect opportunity arose for me to leave Intel and join Zan Compute full-time
What made you interested in IoT?
I joined Intel right when the company was building its Pentium dual-pipeline processor, which means that the compilers were smart enough to run multiple programs in parallel. I was on the compiler team, and we were on the front lines of innovation.
Eventually, I moved to an open source technology team. As part of this team, we started looking at small sensors and small computers, which are the core of IoT, if you think about it. We were working with drones, which are in essence, connected devices. We worked with various types of sensors that make it possible to view many different things, including 3D images through 3D lenses.
As we were working on all of these IoT efforts, something that kept bugging me was that everyone was focused on creating frameworks. What about the applications that use these frameworks that make a real impact on peoples’ lives?
A drone delivery system is a fantastic application — especially in places where you cannot reach like war zones and areas affected by natural disasters.
I saw a lot of innovation happening in terms of edge devices. But something kept bothering me in my mind. I couldn’t help but ask myself, “Hey, how do we use this to make a difference in peoples’ lives? What’s the use of this technology?”
Technology for technology’s sake doesn’t make sense to me. That’s when I started thinking, “What are some applications that I can work on?” The timing was right for Zan Compute.
How did the idea come to life, following its initial stages
We paid very close attention to the mundane. Not many technologists think about how things get cleaned or who does the work. We were fascinated by problems that others have ignored. Cleaning isn’t glamorous. But it’s important.
Our office at Intel was in a building with six floors. There were one or two day porters assigned to every couple of floors. We’d see them at a specific time every day, following the same route to pick up items and clean rooms. We became curious.
Junaith spent some time getting to know the janitors in our building, understanding their routes. What he learned is that there were many inefficiencies that were making their jobs harder — including frequent need for unnecessary trips and demand to hire more staff.
The janitorial industry is centuries old. Very few technologists have looked for opportunities to optimize the resources and service. Maybe paper companies have thought about using sensors to replace paper stock. But think about this angle — it’s more about inventory management. Our angle was different. The question we asked was, “how do we provide better service and make it more efficient?”
That’s how the whole thing started. Janitorial services is a $100B+ market with value predominantly locked up in labor.
Beyond the financial opportunity, the market is a promising one. We can change the way that things get done. We’ve created a data-driven approach to optimizing cleaning resources rather than requiring janitorial services staff to operate on a fixed schedule. A defined process, driven by an AI system and recommendation engine, could make everyone’s lives better.
We’ve also been fortunate to find technically-savvy partners in the industry. We pulled in leadership from ABM, the largest janitorial service provider in the United States. They shared advice for how to break into this industry.
From there, we received an introduction to an amazing person who grew up from being a janitor to being the top strategy person at her company, C&W. She understood the process, loved what we were trying to accomplish, and bought into our technology. She said, “Let’s see how to make this happen — let’s talk about efficiency, savings, and so forth.”
That’s how we started.
What did the company’s technical evolution look like?
We started with HVAC sensors but quickly realized that our approach wouldn’t fly for a few reasons: The battery was too big; WiFi isn’t universally available; HVAC is a flooded market. So we shifted our approach to cleaning and focused on Bluetooth devices where we could get a small coin cell battery to last five years.
We also brought in more on-the-ground experts to get a perspective beyond what we could see as technologists and businesspeople.
We wanted extremely small devices, so we changed the design. We made each unit smaller. We brought in Bluetooth gateways and then realized that we needed an LTE connection. In a lot of places, land and WiFi connections are inaccessible, so the best way for the sensors to communicate and share data is through the cloud. That’s where all the analysis takes place.
As you know, from a COVID-19 perspective, it’s very important to say, “Hey, this restaurant already has ten people, and you cannot add any more if you want to guarantee social distance.”
We can also show population density on a heatmap of where people are, where we should focus, and where the usage is high. Our latest sensor, ZanWave, is radar based and can track all activities of people. It can track whether you are sitting at your table, if you have used your terminal, or if you have touched your keyboard. It can tell you when it’s time to clean a room.
When you clean a room, there are about 40 or 50 steps that need to get done. There need to be one or two people inspecting the room and making sure that all of the steps get completed. There are many requirements necessary for compliance. We can help collect data and validate that everything is done.
What that means from a data and communications perspective is that ZanWave generates a lot of point cloud data. If we want to know whether the janitor cleaned the floor or not, we can do that. We can potentially see whether the mopping got done. We can monitor everything from a health perspective. Without Zan, the process is very manual. Zan uses machine learning and sensor data to make sure that everything is running according to plan.
Every Janitor’s identity gets anonymized, so there isn’t an invasion of privacy. There are no cameras that have the potential to make staff uncomfortable. ZanWave technology is radar based, and our sensors don’t use any imaging technologies.
What are some challenges that you’re currently working through?
When we start expanding the data collected by sensors like ZanWave, it doesn’t make sense for us to transfer all of our data to the cloud. Zan would lose its cost efficiency due to the cost of cellular communication. Imagine the number of bytes that would need to be transferred because of the need to do machine learning on the back end. So we will be doing edge processing and thinking about how we compress and encrypt files on the fly.
Since ours is an open platform, we can integrate data from other sensors and technology. For example, we can integrate data from cleaning robots. We can also drive robots using data from our sensors. We can communicate with the autonomous robot to say, “Hey, go clean there.”
So how should we route the data? It might have to go to the cloud and then go back to the robot unless there’s Near Field Communication (NFC) in place. The kinds of innovations that are possible are going to be big.
We anticipate that COVID-19 will give rise to a variety of data based cleaning solutions. Sensors and robots need to start communicating, which will incorporate both a machine learning and sensor data perspective.
We realized some of the monitors we use were communicating more than was necessary for our solution to work. So we put IP blocks in using Soracom Peek. It’s valuable to identify these potential roadblocks early. It will bring our cost down substantially to reduce the number of unnecessary communications.
We have a lot of questions that come up. How do we enable communication through Zan Wave? How do we secure and also protect customer requirements?
How have COVID-19 and the global crisis affected your business?
We expect that our technology will have a role in alleviating the impact of the pandemic. It’s human nature to be together. Even though many commercial spaces remain closed, they will eventually reopen.
Zan Compute makes it possible to guarantee social distancing and a safe environment. We also help facilities managers maintain optimal cleanliness and sanitation while de-risking the impact to human health.
At Boston Logan Airport, we have over 100 restrooms covered by our sensors and our feedback displays. Before COVID-19, they noticed a decrease in overall quality complaints and were also able to increase their speed for complaint resolution. A stadium in Boston also noticed a decrease in complaints.
Here in the Bay Area, a campus started with a small pilot of about 3,000 sensors and noticed a 40% reduction in cost. They realized that the sensors pay for themselves within a couple of months. With the savings coming from one building, they were able to pay for upgrades to additional buildings.
During the global COVID-19 crisis, we are already seeing an increase of people using occupancy alerts and cleaning cycle updates.
COVID-19 has had a big impact for us as well as you can imagine because all the commercial facilities are closed. They’re slowly starting to open.
I think post-COVID, we will come out stronger than before, but it is a tough path to cross. We are extremely optimistic about the future. We know that we actively need to build it, as entrepreneurs. We can do it. There’s no other choice.
Check out the case study to learn how learn how Zan Compute is building AI-powered sanitation tech.