Optimizing Facility Management with IoT and Autodesk Tandem

VIJAYA KRISHNA ADIGOPULA: Hi, everyone. My name is Vijaya Adigopula, and today I will be talking about Autodesk Tandem. The topic name is, basically, Optimizing Facility Management with IoT. And here is the safe harbor statement. And today, I will be talking about, what are the advantages of using IoT in the facility management domain using predictive maintenance, basically?

So in this session, I will be talking about, what is IoT? What is digital twin? What is Tandem? What are the key features in Tandem? How are you going to classify the assets, how we can create the asset classification? And also, how are you going to extend the parameters, custom attributes, into the [INAUDIBLE] model?

And we will be talking about how we can bring the real-time streaming data from the physical systems, like conditional parameters, using IoT. And then we'll be talking about the predictive maintenance and analytics. And how are we going to reduce the maintenance cost for the facilities management using Tandem application? And we'll be talking about future trends, conclusion, and question and answers, finally.

Yeah, about myself, actually, my name is Vijaya Adigopula, and I have around 24 years of experience into architectural engineering and construction domain. So prior to this, WSP. Currently, I am working for a company called WSP. Prior to WSP, I have been-- I've worked for the companies like Atkins, Accenture, Autodesk, NEi Nastran, CT Core Technologies, Tata Consultancy Services, Satyam Computers, and CADCAM-E. These are the companies I have been working for, multiple, like aerospace, automotive, engineering, construction domain.

So I was into automation, basically, working on using cloud services, like Python, C#, C++, JavaScript, PowerShell. These are the technologies I've been working for, like engineering automation and developing some custom apps for ACC, Architectural Construction Cloud, and ProjectWise, and other Autodesk [INAUDIBLE] applications, basically. I did my master's in MTech, Machine Design from Visvesvaraya Technological University in Bengaluru. And I did my Bachelor's in Mechanical Engineering from Nagarjuna University.

And first, I will be talking about, what are the current challenges which facilities management currently facing? And then we'll be talking about, what is IoT? And what is digital twin, basically? So what are the current challenges for the facilities management are they facing at the moment? So there is a lack of real-time data, and inefficient maintenance process, and difficulty in maintenance management systems, like complex systems, like HVAC systems, space utilization, high-operational cost.

So how are we going to reduce, actually, the maintenance cost, or asset downtime, or high-energy consumption, or the space utilization, like using occupancy patterns? How are we going to do that using predictive maintenance? So this is the whole idea. This session, I will be talking about, how are we going to get the model single source of truth, like ACC, Autodesk Construction Cloud, visualize the model, get the real-time data, and do the predictive maintenance, like get the actionable insights? And how are we going to make decisions out of it?

So if you go to IoT, what everyone is talking about IoT, actually, it's a niche technology. So what is IoT, basically? So IoT is like a network of interconnected devices embedded with the sensors and softwares, basically, enabling them to collect and exchange the data over the internet using 5G. So the role of IoT in facility management, basically, remote monitoring of assets.

Suppose if you have a building, actually, multistory building. So if you want to monitor all the assets, like critical assets, like maintainable assets, take an HVAC system. If you take-- in the HVAC system, you have chiller, compressor or heat exchanger, condenser. How are we going to monitor, actually? So all these assets, like even space utilization, occupancy, lighting systems, so we can use the IoT with the real-time data and we can monitor, actually.

And also, if you look into predictive maintenance, basically, like using machine learning and [INAUDIBLE], utilize the sensor data to anticipate, address, the maintenance needs before any asset failure occurs, basically. And also, you look into energy management. How are you going to monitor and controlling energy usage and reduce or optimize the energy consumption in the real-time environment?

And if you look into space utilization, like occupancy, analyze the occupancy patterns to optimize space allocation and planning, and enhance the security and safety. So these are the key features in internet of things. And now, if you come to digital twin or the asset lifecycle management, so what is meant by digital twin, basically? So if you look into digital twin, so digital twin is like a virtual representation of a physical asset.

So we are going to ingest multiple sources, like data from CRM, like CMMS, computer maintenance management system, computer-aided facilities management, CAD data, BIM data, GS data. Everything will be ingested into cloud environment. And visualize the data. And from out of it, getting the actionable insights and how the end users make decisions out of it, like how the facilities management.

Once we handle the building model, like asset information model, to the cloud-- facilities management, how are they going to remotely monitor all the critical assets or maintainable assets remotely? And what are the actions integrate with the CMMS system, like IBM Maximo, Jira, available in the market, basically? They're the key applications.

So automatically, how are we going to-- based on the anomaly, with the real-time data, we can anomaly or asset failure prediction. Based on that data, how are we going to monitor and take the decisions, like what type of maintenance activity need to be performed for the technician, automatically assigned to the technician, create a work order? Everything, how we can be automated, I will be talking about.

So what are the digital twins I spoke about? And then application of digital twins in facilities management. So it's like creating the virtual representation of a 3D model. So connecting to single source of truth, like any Autodesk Construction Cloud environment or from [INAUDIBLE] native system, basically, import the models. Currently, the digital twin application, for example, if you take Tandem, Tandem supports IFC, Revit, Navisworks, DWG file formats.

So we can import the model, basically, and create a virtual representation, connect to ACC environment, get the model, and create the facilities template, and classify the assets. And then extend the attributes if you want to create custom attributes. And then perform the predictive maintenance based on the historical data.

From the physical system, like in the billing system, like HVAC system, if you look into these HVAC system, they are continuously sending the conditional parameters, like energy consumption, airflow, temperature, pressure, humidity, vibration. These kind of sensor data will be getting into the cloud environment.

And using any kind of Azure services or AWS, get the-- develop some custom triggers. How are we going to read the data from IoT hub, and ingest it into a digital twin application, and perform the predictive analytics, and get the valuable insights, make the decisions out of it, like what type of maintenance activity we need to be performed? And how are we going to optimize the energy consumption space utilization in the building, basically?

So we can monitor and do performance monitoring, we can do the track, analyze asset performance, how it is performing. And also, we can look into asset [INAUDIBLE] health, inventory health. A lot of things, we can perform using asset lifecycle management. So this is like, if we look into what is digital twin, it's like a design, construct, operate, and also decommissioning of the 3D model, real asset, basically, so managing entire lifecycle of a building from design, construct, operate, and decommissioning.

And these are the maturity levels for a digital twin application. So if you look into Tandem, they have started in 2021. And 2023, they have started in facility monitoring and predictive maintenance, like with the real-time data. They have started integrating the sensor data, connected. They had given flexibility, how we can get the real-time data from Azure Cloud Services. And now they are working on comprehensive and autonomous. Autonomous is the high level, actually. So right now we are in the comprehensive stage.

And now we can look into the Tandem. So Tandem is a digital twin application. Basically, we'll be hand over to that facilities management, along with the asset information model. So we'll be having-- it's like a consolidate data from different sources, like IoT devices, building management system, enterprise ERP data, everything into a unified platform, basically, for comprehensive facility management application, basically. It's a data aggregation.

So it's like a centralized platform, like [INAUDIBLE] track, manage facility assets throughout the lifecycle, basically, installation to maintaining and decommissioning, and also the visualization. That is the good part of the-- Tandem is very good visualization. You can import, like I said, Navisworks, IFC, Revit, DWG files into Tandem application, basically.

And also, we can perform that clustering. We can classify the assets, and we can-- at a time, you can select based on the classification. And we can assign, actually, a unique classification code or master template. Or real estate code template, we can assign to the classification system. And then we can perform that predictive maintenance, basically, used on that classifications. Or the critical assets are maintainable assets, basically.

And integration with other systems like Tandem seamlessly integrated with other CMMS systems, like IBM Maximo or a Jira application, other things, actually. And like I said, data analytics. So that's a good part. We can perform the data analytics with the historical data, time series historical data. And we can find the anomalies, like spikes in energy consumption or temperature. And we can optimize the operations and make decisions out of it.

And collaboration. Tandem facilitate collaboration between stakeholders, enabling them to access, share, and interact with facility data and models informed for model-- informed decision making and security. Here is the high-level architecture for the asset lifecycle management Tandem. So from Tandem, actually, like I said, it's a comprehensive application. Basically, it's like a unified platform, basically.

So in Tandem, what we'll be doing? First, we'll be creating the new facility template. We're importing the Revit model, and we were classifying the assets. And then that's getting all the asset information data from the [INAUDIBLE] 500 model. So you can warranty sequence number, everything like asset-related information, everything we'll be getting from the interface, Tandem interface.

And also, with the real-time data, we'll be integrating with the Tandem, like using data connector. They have implemented one application called Tandem Connect. So using that, using BMS data or CMMS, everything can be integrated and ingested into a Tandem application. So we have two SR twins, basically. One is asset twin with the [INAUDIBLE] 500 model data. And the other one is performance team with the real-time streaming data, so where you can see the visualization data, everything.

And if you look into a facility management with the Autodesk Tandem, these are the key components, like monitor, investigate, and act. It's like a holistic view of the building performance of the asset. So all the assets, we'll be monitoring all, like I said, HVAC system. We can monitor heat exchanger, compressor, condenser, and chiller components for any kind of space utilization, anything we can do, like remote monitoring.

And investigate. So once we got the anomaly, we can identify-- based on the trends and patterns, we can see what kind of analysis and root cause of the failure prediction, basically. So we can look into failure components. And also, we can take-- instead of going for reactive approach, we can convert reactive to proactive and fix the components before the failure, basically. So that is the whole idea, to reduce the maintenance cost, and also optimize the energy consumption, reduce the maintenance schedule, everything, actually, with the use of Autodesk Tandem.

And like I said, here is the architecture for the Tandem applications, where you can see what kind of data we can ingest, like a SCADA/IoT, computer maintenance management system, like I said, IBM Maximo, ERP application, like a JD Edwards, any kind of where they are tracking for work order information, equipment master related to top historical data for the maintenance of the asset data, like everything, like how many times the asset went for the maintenance activities, like what type of maintenance.

Is it corrective maintenance, preventive maintenance, major maintenance? So everything we can ingest, actually, into this Tandem application. And also, BIM, we can import the ACC environment. It's seamlessly integrated with the ACC environment. So what are the files we can import into Tandem? And we can get actionable insights out of it and make the decisions, like how the facilities management getting benefited out of it, like workshop operator, workshop manager, maintenance coordinator, or facilities manager.

And if you look into how the real-time data is integrated with the Tandem connector, so Tandem connector is a very new feature in Autodesk Tandem. It's in beta stage at the moment, but they have very good features. Like I said, we can integrate CMMS application, Maximo, or BMS data using BACnet or MQTT sensor data.

So we can easily-- it's like a low-code event-driven application, so where you can easily integrate all the temperature, humidity, like proximity, touch sensor, motion sensor, whatever, pressure sensor, gravity. So these kind of sensors, we can integrate with Autodesk Tandem using Tandem Connect. And we can perform the predictive maintenance, improvement of the operations and maintenance based on the threshold values.

So any anomaly breaches, we can look into maintenance activities, basically. It's like a regular-- a preventive maintenance, like look for engine oil change or filter change. Any cleanup activities required for the assets, basically, all these we can look into. And also, automatically, it will send notification to technician and create a work order, look for any spikes in energy consumption or temperature, basically. All these things, we can monitor and use the Tandem application for investigate and reduce the maintenance cost.

So this is actually like, how are we going to integrate different data streams with Azure SQL, with the Tandem Connect, actually, with the CMMS data or real-time streaming data? And how are we going to store the historical time series data in Azure SQL? And how are we going to perform the predictive maintenance? And what are the actionable insights and the key features we will be evaluating? And how are we going to build the dashboard?

And what are the types of-- suppose if you are developing any kind of a digital twin. So our aim is like, what I'm going to achieve? What are my strategic objectives, goals, benefits, outcome I'm going to visualize? So we have to come up with all the use cases and scenarios like, what are my-- how are we going to reduce the capital expenditure or operational expenditure, basically?

So if you come up with all the use cases like, I'm going to monitor asset performance, or asset health, or inventory health, or I want to see how the asset is performing, like mean time between failure or mean time to repair, asset reliability, all these parameters, we can calculate on the back end and we can visualize in the dashboard, basically. That is the advantage.

We're integrating with all the sensor data, machine learning models, real-time monitoring of all the assets related to building, combine all of them and get all the insights into the dashboard. That is called KPI, key performance indicators, like everything we can see, red, amber, green. Is it like asset is performing in the right condition? Or everything can be monitored using Autodesk Tandem.

So in the left-hand side, you see the diagram. We will be ingesting into ACC BIM model, like IFC, Revit model, like standard operating procedure documents. We can link the documents. We can import the files from different source of files, IFC, Revit, or DWG, Navisworks, CMMS historical data or Azure sensor data, everything ingested into Tandem. And also, you can ingest into Azure SQL for the historical data. And we can perform the simulations.

And based on the simulations and interventions, we can decide the component. We can go for the maintenance activity now or we can delay for the maintenance activity, all these interventions, like what type of maintenance activity need to be performed? Based on the anomaly asset failure prediction, we can do all those interventions, basically.

So at the output of the machine learning model, the predictive maintenance gives a mean time to repair between failure, inventory health, prediction, asset reliability, all those things. We can forecast everything, actually. And we can get these insights into dashboard. So that has been integrated with the Tandem. So this is the entire workflow, like how that real-time data and different sources, like structured, unstructured, semi-structured data ingested into cloud, and bring all the-- perform the predictive maintenance, get the valuable insights into Tandem application.

So this is the beautiful feature in Tandem, where you can see a clustering of all the building, the entities, like building elements, like categories. You can see duct elements, air handling units, furniture, mechanical equipment, pipe fittings. You can see all are categorized, actually. So using the-- there is a drag option here.

You can select all group of entities, like pipe fittings or air terminals, with the single selection and classify those assets based on the uniclass classification code, or real estate code template, or master template, whatever provided by Tandem. And once we classify the assets, we can easily ingest the sensor data, real-time data, and we can perform the predictive maintenance, basically.

And where you can see the dashboard is a very good visualization. With the one click, you can get everything into one dashboard. And also, right side, you can see the inventory health of the asset, basically, like asset tagging, where the component belongs to any one of the asset, like which floor, and which level, and which room that asset belongs to.

So everything, we can get everything related to inventory and the asset tagging, labeling of the assets, everything into a visualization of inventory or panel information, where you can see on the right-hand side. And below you can see the access control, so where you can see assign-- add users and set the permissions, like read and write permissions, for this particular Tandem application. So all these features are feasible.

And also, you can see the filters, like how you can filter the model, like Revit model, based on the levels, or spaces, or classification assembly code, or Revit category, whatever I said, based on the category, like air handling unit, or mechanical equipment, or duct or fittings, whatever. Based on the category, we can filter and we can visualize the group of entities, basically.

And with the real-time data, we can set the threshold as well. Any anomaly breaches, we can send alerts to technician and facilities management to take the immediate action and see preventive maintenance, like continuously monitoring what is happening for the asset, basically. And also, you can create the custom views, and we can save it into the Tandem application.

And also, we have systems, mechanical system and system tracing. Actually, this is one of the very good feature. And also, dashboarding, I said, very good visualization. Quickly, we can create on-fly for this dashboard. Yeah, so these are the very good features in Tandem, like I spoke about so far. I have a-- what are the challenges facilities management facing? What are the good features in Tandem? How we can get the data from different sources into Tandem application, and perform those actionable insights, and make decisions out of it?

Now I will go with the one real use case, like HVAC system, how we can do-- perform the HVAC system predictive maintenance? And what are the types of maintenance? If I run the machine learning algorithms, like for anomaly and asset failure prediction, if I run those two, and I want to see the output, like how the data looks like for these kind of HVAC system, like chiller, heat exchanger, compressor, condenser, or get this data and what type of maintenance activities need to be performed.

And based out of, how am I going to reduce the energy consumption of the asset? And also, how I'm going to reduce the energy consumption, as well as the maintenance cost of the asset and maintenance schedule? All these, how I'm going to perform, basically. So this is the use case I'm going to explain now.

So the main use case for the requirement, basically, in the HVAC system, the whole idea of my requirement is basically like how I'm going to improve the reliability of the asset to perform by using predictive maintenance, machine learning algorithm, ML, model, and identify trends and patterns, and also the real-time streaming data and historical maintenance records, and how I'm going to optimize the energy maintenance and maintenance schedule, and increase the life of the-- improve the life of the asset, basically.

So the key components monitor, basically, HVAC system related to chiller, condenser, compressor, and heat exchanger. And the sensors have used energy consumption, airflow, temperature, humidity, vibration to collect the real-time data and predictive analytics. So where you can see here, I have ingested real-time data into Azure Cloud Service, where you can see I have created four devices. You can see condenser_001, chiller, compressor, and heat exchanger.

So the IoT hub, this is like entry point for like Azure Cloud Services for the real-time data, time series data, we call it as. So how we can continuously sending the real-time data from device to cloud? So once you have the data in the cloud, actually, so there are a few connectors, cloud connectors. Suppose I have a motion sensor, like an airflow sensor, like humidity, temperature.

So we can use-- there is a Cloud Connector with one connector. Everything will be ingested into Azure IoT hub. And from there, using event grid, trigger function app, we can ingest all the data from IoT hub to Tandem application, basically. So after ingesting, actually, you can see that the green symbol, the live data has been connected.

And you can see that device has, basically, chiller components related to HVAC system, compressor, chiller, condenser, and heat exchanger. And below the right-hand side, you can see the live sensor data. So when user clicks on any kind of bubble, actually, you can see the sensor data belongs to that particular component.

And this is how the CVS dump actually-- historical data for August month I have taken for analysis and predictive maintenance. And I want to see how I can optimize that energy consumption, basically. So here you can see timestamp with the component name, energy consumption, airflow, temperature, humidity, and vibration.

So I have developed two machine learning models, basically, one for anomaly detection and second one is asset failure prediction. So anomaly detected, there are two scenarios. Suppose if the anomaly detected-- the machine learning model gives true or false. Based on that, if it identifies any anomaly detected, then we have to perform that conditional-based monitoring, like I said, basic maintenance activities we need to perform.

And based on the probability of the failure, actually, like a performance-- asset failure prediction, probability score, it will be returning the second one. So I have used two machine learning models. One is called Autoencoder. That is for anomaly detection. And second one is called XGBoost. This is giving both unsupervised and supervised learning, basically, finding-- analyzing the patterns and trends, and give the model output as two things.

One is a high probability of asset failure, a score. And the second one is we need to calculate the remaining useful life of the asset. So there is a formula, like 100 minus the percentage of probability of asset failure. It gives the remaining useful of the component. So if the short of-- the remaining useful life, that means it's like an immediate-- corrective maintenance action is required. Intervention is required, actually.

So we have to look for replace or repair the component. So suppose, if it's a compressor, we need to replace the bearings. So like that, we have to validate it. If it is like an anomaly detected, we have to look for the regular checkup, basically, like it's preventive maintenance. If the remaining useful life is between 20 to 60, we have to plan for major maintenance, like a preventive or major maintenance.

And if it's a long ordeal, it's greater than 60, that means it's like a routine monitoring, basically. So low remaining useful life means high-failure probability, basically. So both are inverse. High probability is equal to 1 by RUL. So based on that low-- we have to look for anomaly detection and asset failure. And we have to take the maintenance activities. We need to perform corrective maintenance, or preventive maintenance, or major maintenance.

So we have high anomaly detection and plus high-failure probability, so we have to go for immediate intervention. So we have to either replace the component or a complete shutdown. And we have to overhaul the entire component, basically. So these are the action items based on the predictive maintenance ML model output. We have to make the decisions. Maintenance team need to make the decision, basically.

So like I said, the corrective maintenance, if the maintenance type is corrective, then we have to go for-- this is triggered by high-failure probability and low remaining useful life of the component. If it's a preventive maintenance, like it's a moderate RUL, remaining is life and recurring anomalies. If it is a conditional-based monitoring, like detect anomalies, but low failure probability. If it is like a major maintenance, like severe issues, critical systems are overhauling requirements, basically. So this is very important for the maintenance activities based on the predictive maintenance ML model output.

And you can see, this is how ML model outputs. It will output anamoly score, like anomaly, like failure prediction. You can see the failure prediction. And from there, you can see that the remaining useful life is calculated. So in this case, remaining useful life is very high. That means the asset failure, you can see that anomaly and asset failure prediction.

So high means it's like, no need to worry about either corrective or major maintenance. So it's like, just no immediate action is required, just like continuous monitoring, like a regular checkup and replace the filters or this basic checkup need to be looked into. So like that, these are the two scores, anomaly score, and failure prediction and remaining useful life. Based on that, we need to perform the maintenance activities.

So how the maintenance activities based on that, how the system has stored, actually, in the CMMS, computer maintenance management system, like IBM Maximo or Jira. So with the timestamp and work order, it will be automatically created. And you can see that intervention, what intervention need to be taken care, and remaining useful life, and which technician has been assigned, and the maintenance cost for that particular asset. And what are the activities performed on the particular asset, basically, with the maintenance activity?

So if you look into preventive, like I said, preventive is kind of a conditional monitoring, like lubricating, like changing oil filters, basic kind of maintenance activities. And if it's a major maintenance, it's like [INAUDIBLE] is less than 20. So there you can see all 15.8, 10.45. It's like a complete shutdown and replacing components, major components, like related to HVAC compressor or related-- or condenser, heat exchanger, or chiller kind of components, basically.

So anomalies are detected. Like I said, these are the components, like detected sudden spike in energy consumption or vibration in the compressor, indicating potential bearing or refrigerant issues. And if it is a chiller, like temperature fluctuations, observed by significant temperature variations in the chiller, suggesting potential heat exchanger fouling or coolant problems, basically.

So condenser boiler, like a coil blockage, like a flagged gradual increase in condenser energy usage are leading to a discovery of dirt and debris buildup on the coils, basically. And also, if you look into heat exchanger efficiency drop, basically, detected decrease in heat exchanger efficiency, prompting an investigation into potential falling leaks.

So these are the operations and maintenance based on the optimization and interventions need to be taken care, so monitoring real-time data, predictive maintenance to prevent overuse, like automated optimization, anomaly detection, recommended interventions based on the output from ML models. So this is the architecture, like how the predictive maintenance is carried out, actually, from sensors, like sending the real-time data.

So Azure hub, actually, like this entry point, where the time series data has been ingested. And from predictive maintenance, using ML models, like I said, XGBoost or Autoencoder for the unsupervised and supervised learning models. Perform those machine learning models, and get the output, and see the remaining useful life. And based on that, perform those actions in Tandem. Visualize those anomalies and failures, recommend for the interventions and facilities management dashboard, and providing AI-powered insights for the preventive and corrective actions.

And if you look into the last one, like CMMS, it has been tightly integrated with Jira application. And very soon, we'll be seeing that integration and release it in market from Tandem, actually. So this is like a predictive maintenance workflow, like data collection from HVAC system example, so like energy consumption, temperature, airflow, vibration, sending sensor data to cloud, and then anomaly detection and then failure prediction.

And from there, we'll be calculating remaining useful life, maintenance recommendations, and manage facility management, how they are getting benefited with the maintenance activities. So these are the libraries I have used, like Autoencoder for anomaly detection and XGBoost for failure prediction. And from there, I'll be calculating remaining useful life estimation out of it.

So these are the maintenance recommendation based on the ML model, like a compressor chiller, condenser heat exchanger, preventive maintenance, and energy optimization. So you can see that plot for anomaly and asset failure prediction for all the four components, the spikes in the energy and temperature, basically.

And from there, with all these data, we'll be calculating asset health or reliability. How are we going to improve the asset reliability? Total uptime-downtime interventions, like recommendations, remaining useful life of the component, and asset failure prediction. So everything will be calculated, like mean time between failures, mean time to repair. For all the asset information related, everything will be visualized.

So overall, if you perform those maintenance activities, like overall energy consumption has been reduced to 15%, system downtime is reduced by 20, and maintenance cost saved by 30%, basically. So the whole exercise for this proper, like how are we going to ingest the data into Tandem application, and ingest the real-time sensor data, and store it in Azure SQL, and perform the machine learning-- I mean predictive maintenance, and get the insights, like perform-- calculate the remaining useful life of the asset and make the decisions out of it?

How are we going to reduce the asset downtime and increase the reliability of the asset, improve the reliability of the asset and predictive maintenance, extend lifespan, and reduce the overall maintenance cost by 15% to 20%, and also the scalability and integration? So this is the example.

We did it for an HVAC system, but the same thing can be carried out for other critical assets or maintainable assets in the building model, basically. So how are we going to ingest the data? How are we going to build the platform? Where are we going to store the historical time series data in the cloud services? How are we going to integrate CMMS data?

How are we going to perform that inventory health, asset health, asset performance, remotely monitoring what type of assets based on the use case requirement for the facilities management, basically? So we have to build the ML models. And how are we going to show these KPIs in the dashboard in Tandem? So these are the key things for the facilities management. And they can take the-- from the actionable insights, they can make decisions out of it, basically.

So final thoughts, basically, like optimizing facility management with IoT transforms, how the organizations may manage their assets, or related to HVAC, or space utilization, or lightning, or anything related to security, offering predictive insights, reducing downtime, optimizing energy consumption. This combination of real-time data with predictive analytics and AI-driven decision making, and powers facility managers to achieve operational efficiency and long-term sustainability goals, basically.