The Covid-19 pandemic is fundamentally altering our thinking on how we collaborate and cohabitate. As the social distancing regimen extends from days to weeks to months, many people are getting attuned to the idea of doing almost everything from the safety of their homes. In the few situations when they do venture out, they are keeping their distance from others and touching only what must be touched, ideally after a liberal swab down with disinfecting wipes. Masks have now become standard attire in public spaces, and it requires some amount of mental strain to recall what things were like just three months ago.
This lockdown is going to have long term effects. When it does inevitably come to an end, we will likely never get back to the levels of free movement and inter-mingling that we took for granted in the past. Instead, we will probably see the emergence of a new world where many interactions are managed and curated. As contact tracing gets implemented and becomes a normal way of life, humanity could find itself clustered into red, yellow and green categories, with associated privileges and access levels. Appropriate distances will need to be maintained in public spaces and the environment will need to be kept clean and sanitized.
The places where a lot of this will play out is in public buildings like offices, schools, and transport hubs. These buildings by their very nature tend have a high density of people operating in close proximity and physically interacting with the infrastructure. We will need to deploy all sorts of technologies and processes to make our interactions in such spaces safe and convenient, while hopefully retaining some level of amity and warmth that are so essential to the human experience.
We will likely see contact tracing data being coupled with movement tracking and access control to manage social distances and prevent mixing between people with different classifications. Wayfinding applications may now have an additional layer to create personalized routes and quarantine areas. While we will see more frequent cleaning, we will likely also see many physical interactions with hardware get replaced by touchless controls. Technologies like image recognition, wireless sensing and cloud networking will likely play a strong role in making all this happen.
According to a report from Gartner last year, building automation was expected to be the fastest growing IoT sector in 2020 with an estimated installed base of half a billion end points. Given our new realities, this trend will likely accelerate and potentially dominate new investments in IoT and M2M technologies in the coming years. While energy usage and asset management have been the key drivers for building automation in the past, we will likely see more investment in use cases related to people flow management and facility sanitation as we go forward.
Decentralized Control Can Truly Open Up the Smart Building Ecosystem
Building automation is not a new field. Over the years, a lot of work has been done by a number of well-known system integrators, like Honeywell, Johnson Controls, Schneider Electric and Automated Logic to name a few, to integrate various building systems such as HVAC, lighting, access control and fire suppression. The typical mode of operation is to have all the data from these assets come via a communication network to a central server that hosts a “Building Management System (BMS)”, which processes the information, makes decisions either autonomously or with a human in the loop, and sends out actions in the form of commands. This sequence of sensing –> decision making –> actuation is used to orchestrate and automate all sorts of outcomes such as energy savings and building occupant convenience.
While the above approach has worked reasonably well, it is highly dependent on the BMS which acts as a central brain and is in effect the single point of decision-making. Every sub-system or application in the building usually sits within its own private silo with all data aggregation and cross-system decision making left to the BMS. This approach can add a lot of complexity even for a simple action like switching on the lights in an office when the occupant arrives in the hallway. As requirements become more complex and the use cases grow, a more efficient approach might be to decentralize the decision making as much as possible and only use the BMS for executive decisions. This type of empowerment paradigm where most decisions and actions are executed through coordination at the edge could really open up the landscape. It could enable all sorts of innovative new subsystems and applications to easily join the building network, readily exchange data within local neighborhoods and clusters, and enable intelligent outcomes.
In large enterprises, decentralization would also promote variable feature implementations depending on the specific needs in a building zone, making things more cost effective for both the owner as well as the tenant. Unsupervised learning at local and global scales could be used to identify patterns, which could then be used to implement certain technologies and features only where they make a big difference.
Finally, a decentralized approach would be inherently more resilient to external disruptions because the brains of the operation would now be distributed all over the enterprise, rather than at one single point of failure within the BMS. In situations that demand coordination e.g. emergency evacuations, the various clusters could coordinate with each other in real time to assess current state and drive optimal decisions.
It seems evident at this point that decentralized automation has a number of merits. What do we need to make it happen?
Building Communication Needs a Narrow Waist
Adding a new feature into an existing building automation system is typically not easy. If a vendor comes along and offers a solution to monitor something new, e.g. leaks in your roof, they will likely offer a solution that uses their proprietary sensors that send data to their proprietary server or cloud, where insights are gleaned and then sent to you either via an API or through more conventional tools like email and text. The job of integrating this information into a building-level intelligent system is up to you, the building manager. And herein lies the problem. If you wish to share the data with other building systems, then it is not straightforward for the involved parties to perform a handshake and proceed. Instead, they usually need to go through a lengthy integration exercise with each other, and even once this is done, every data exchange instance can involve navigating many layers of protocols and networks to get information across from one node to the other.
Sounds painful, right?
This is a major reason why we have not seen the smart building ecosystem grow rapidly. If you look at the integrated solutions being provided by the large manufacturers mentioned earlier, all of them tend to provide solutions which leverage their own sub-brands, coupled with their own proprietary BMS. They may offer to integrate with solutions from other companies but with the caveat that there will be a cost, it will likely be time-consuming, and there would be a risk of experiencing glitches. This is because everyone uses somewhat different communication protocols and every integration exercise requires some special interfacing software to be written and tested. While these are all valid reasons, they unfortunately make it almost impossible for building owners to mix and match the best solutions from different providers. If we had a universal protocol for communication, we could remove these barriers and promote innovation.
Is it possible to create a universal protocol for building IoT? There is a precedent for this. The growth of the internet was enabled by what is called the “narrow waist” of communication and it may be an approach that could be leveraged in building systems as well. To understand this concept, please look at the below figure. The way the internet works, all devices must communicate using the internet protocol or IP, which assigns a unique address to each node and ensures that any message addressed to it, finds a way to reach it. This network layer is the narrow waist that all internet traffic must go through. However, there is a high amount of flexibility both above and below it. Devices can still sit on their own local networks and communicate via a variety of physical media and protocols within their sub-nets. Additionally, once messages are delivered, there is huge flexibility for applications to do whatever they want with the information.
This unifying narrow waist structure is what enabled the large-scale connectivity and wide range of applications on the internet that we take for granted today. If we could implement a similar universal narrow waist for building communication, it would democratize access and unlock a lot of innovation. It would allow us to easily interconnect new solutions and networks while providing freedom to application developers to build without needing to know the underlying technology. It would be like creating an app store for buildings – anyone could develop solutions as long as they know the rules of the game. It would also enable easy and universal integration of buildings with mobility solutions, last mile delivery solutions and municipal infrastructure solutions, and effectively create a seamless continuum of experiences for their tenants.
What would a universal protocol look like? In smart buildings, a lot of the applications will not require a central command and control unit, but rather an ability to freely and securely exchange data between nodes. A good protocol for this might be Named Data Networking (NDN) which promotes easy data sharing and emphasizes broadcast and consumption in a high trust environment. Unlike the Internet Protocol whose primary focus is end-to-end communication, NDN is focused on distributing data, and fits well with our proposed decentralized approach for building automation.
We Need a Commercial Backbone for M2M Transactions
Now that we have discussed the mechanics of building automation, we need to talk about another important consideration – monetization. In a decentralized world, we would need to have some mechanism to keep track of the value being provided by the various applications. A record would need to be kept for every single time something of value was enabled by any application, as well as every time a piece of data published by one application was consumed by another one. The first transaction would be billed to the end customer whereas the second would be cross-billed to the other vendor, who in effect was a partner in enabling an outcome.
There are companies who have already started working on solutions for managing and billing these types of micro-transactions. They would likely use some form of tokens that are collected over time and turned in at the end of a billing period for reconciliation and payment. Since this would be a model where the customer would only pay for what they consume, there could also be mechanisms to set limits (like a prepaid SIM card) where if the quota for the month is exceeded, the feature would stop working unless the budget was topped up. This is an evolving ecosystem but soon to be a highly profitable one (imagine getting a cut every time an M2M communication occurs in a building!)
A Peek into the Future
We have now discussed three key enablers that could really expand the possibilities in smart buildings. What might life look like in such buildings with decentralized and coordinated automation? Here are three examples to jog your imagination.
Flow orchestration: In a decentralized world, the orchestration of many automated experiences could be handled locally without the intervention of the BMS. When a tenant enters a building and has been authenticated, the mobile credential that they are carrying (e.g. their cell phone) could create a seamless chain of custody and enable a series of actions as various sensors pick up its movement. Calling the elevator, switching on the hallway lights, opening locked doors, turning on the office HVAC system, etc could all be done through local communication and handoff. The exact same approach could be applied to a last mile delivery robot or drone.
Adaptive sanitization: In the post-Covid world, we will likely see the deployment of cleaning robots, UV lights and airborne disinfectants to keep our public buildings clean and safe. All of these actions would need to occur as often as possible, typically after one set of humans have passed through a workspace and before the next set arrives. Unsupervised learning could help with understanding the needs in various building zones, which could then be used to schedule the cleaning actions in a highly efficient manner. Swarms of robots could self-coordinate and deploy based on real-time demand during the day and fall back into a pre-scheduled operation in off hours.
Emergency response: Imagine an emergency situation such as a fire or active shooter in the building. Responding to these requires high situational awareness and there may be no time to bring human decision-makers into the loop. In such scenarios, a decentralized model could be quite powerful where cameras and people trackers could rapidly assess the location of the threat as well as the tenant population, based on which the access control system could partner with the way-finding system to create movement zones and lock-down zones, effectively blocking in the threat and ferrying people to safety.
OK, let’s get back to our present reality. There is clearly a lot of work to be done before we can implement this type of decentralized control in buildings. Specifically, we will need to deploy a universal data-centric communication protocol as well as a new micro-transaction based payment system. The question is – who will take the lead and make it all happen? Will the incumbents do it? Or will it be done by newcomers like Microsoft, Amazon or Google who are also building a variety of cloud based IoT platforms? Or will it be done by a trusted intermediary, such as a standards organization, that coordinates with all the interested parties to build out a common architecture?
The first mover will get to define the new rules of the game. I look forward to participating in the action as it unfolds.