Occupancy Data Driven Facilities Management Frameworks

The FM Market in the UAE is maturing. We are seeing clients move away from manpower based solutions toward a more holistic building functional approach based on human/building interactions.

In saying so, the actual building function should now be the driving force of our maintenance programs and not just a simple PPM program and reactive team running around a building. Technology has advanced to a point whereby key occupancy data is now the future driving force of the FM industry both for asset owners and for FM service providers alike.

With much talk about predictive and conditioned based maintenance frameworks from the service delivery companies, the asset owners are looking to technology companies and FM Consultants like us, to set up the framework and organize the network of smart sensors devices and existing building management systems to deliver occupancy data that can be used to better plan shifts, reduce manpower and increase efficiencies. This blog article will discuss how we do this and what types of savings can be expected.

Occupancy Density Sensors, Artificial Intelligence & Machine Learning Developments for Facilities

The technology field in radar sensing is providing a new pathway toward measuring occupancy without the use of cameras or face detection, shape detection or other potentially invasive solution. The rapid development of Millimeter Wave Technologies that do not detect faces or shapes but rather detects range, velocity, and angulation for multiple objects. These sensors work in no light, low light or even dusty or smoky environments which makes them ideal for a variety of indoor and outdoor applications in facilities from internal office spaces, to dark car parks, to large outdoor sporting stadium venues. The inbuilt analytics is trained to understand the movement. The important point here is to understand that through the use of MMW sensors and in built sensing logic, that massive amounts of occupancy and human behavioral data is collected and stored in a data base that can be used to predict, measure and improve performance of our teams in both the hard and soft aspects of the FM Service delivery.

As an example, we can now understand clearly which parts of a commercial building are used heavily, or where people tend to flow within an airport after a certain flight to be able to better place and plan reactive and spot soft FM services, or where and when to plan invasive and disruptive PPM within an office space sub-ceiling.

Artificial Intelligence Applications with Data

Because these sensors are gathering behavioral data and aggregating this data into one singular place, we as FM service providers or building owners now have the actual human asset interaction data that can be used to better predict the degradation of assets based upon utilization. That's apart of course from the potential energy savings of being able to reduce energy consumption through better understanding of the internal variables like temp, humidity, CO2, Ozone and other gases that of produced and often trapped within internal building environments.

Machine Learning Applications Across Time

The beauty of non invasive sensors that are fixed within key areas throughout the building is that these sensors begin to be able to be trained to recognize human intention and also to provide an additional level of health and safety without being invasive. Because these radar sensors work in low visibility situations and can detect even the smallest amount of motion or movement, they are key tools to be used in

• Fall arrest situations in healthcare facilities bathrooms,

• Fire detected in an area that is unreachable or unmonitorable by traditional CCTV cameras for human location and detection - like bathrooms, changing rooms or other sanitary locations.

• Sensitive Data Areas like Data Centers or Document Archives

By utilizing key sensors within these areas we can detect if a person has fallen, if they are still breathing or if they are stuck in a position and cannot release themselves without help. These applications make the use of these sensors as key applications that can augment a health and safety program within a facility or within a key oil and gas/manufacturing facility where large areas and inanimate equipment and production lines carry high risk for employees.

Multiple configurations of frequency and internal sensor components can increase accuracy and computing power to open further doors for FM providers and building owners to be able improve their building function, reduce energy costs whilst reducing risk to occupants.

Performance FM has a full development team available who are experts in building code and improving Machine Learning within Radar sensing technologies, and we can build any application or integration utilizing the data collected from sensors to solve any occupancy objective being faced across a wide variety of facility types.

Occupancy Driven FM Framework Example

Most buildings within the region are already well equipped with a myriad of third party sensors or BMS systems; however, there still remains a large systems integration gap between all different types of systems including ELV Systems (Access Controls, CCTV, Indoor Air Quality Sensors) with the BMS and the Fire Safety System. We see the usual BMS Operator in a room with a team of security guards who are to monitor screens 24x7 and try to adjust and change the set points to influence the comfort of the building.

The below illustration shows how the sensors track human movement with complete anonymity.

However, with the tracking, velocity and movement patterns within different spaces we have no opened up the doors to be able to understand and predict the interaction of space across time in certain areas based on typical human movements and activities within that space.

Similarly we can easily track anomalies in energy consumption in assets that are linked to specific spaces to identify poor asset condition performance and even begin to predict asset failures before they happen.

When asset performance anomalies in energy, vibration, sound, or temp/RH are detected within a space that is empty or of low utilization we can easily determine that there is either a systematic MEP issue or there is an individual asset problem which has contributed to this potential future failure. Our teams due this via using a simple logic based algorithm based upon manufacturer performance data easily benchmarked into the data set system.

The above image shows a CO2 sensor in a food and beverage area at 13:46, which would naturally coincide with the lunch period. This is just a simple application of occupancy matching with existing sensors within facilities. This data is very powerful as we can easily see the impact on the FAHU operating times to reduce the CO2 levels in that particular room and make data driven estimations on how the occupancy affects the time the FAHU is utilized.

FM teams can then easily identify and investigate and make a correction to the asset before a failure occurs. This transparency and visibility to building performance and occupancy space data gives performance based contracts a new level of visibility and allows FM providers to greatly reduce their reactive maintenance teams whilst still ensuring compliance with the SLA/KPI of the contract.

The following existing sensors can be integrated within a building to collect and aggregate data and make internal building performance assessments:

• Radar Occupancy Sensors installed in open rest areas (Receptions, open plan offices, common areas)

• Meeting Room Density Sensors

• Canteen, Food & Beverage Areas Radar Sensors

• BMS Temp/RH Sensors

• CO2 sensors if not already within the BMS System

• Lux Level Sensors for Meeting Rooms & Offices

Here we can see a sample of how occupancy sensors can be compared via an AI Algorithm to identify any "odd" characteristics in utilization within an intermittently utilized space.

The above image shows HVAC consumption compared with meeting room occupancy across a 3 day period of measurement. We can simply see that when occupancy increases, so does the demand to retain Temperature & humidity points within a given space. Utilizing this simple graph and understanding how occupancy and human behavior drives asset utilization, we can easily predict and estimate the conditional degradation of the asset and then change our PPM plans according to this data. We simply are not maintaining for the sake of appeasing a CAFM system. We are providing maintenance where maintenance is due, just like your car has been working for the past 10+ years.

Naturally, this saves money and provides justifiable proof to both asset owner and FM company alike why manpower calculations and PPM Programs are adjusted in such a manner.

Furthermore, the Asset owner and FM team should work together to understand the building occupancy and utilization data to better adjust and predict their asset lifecycle and long term financial planning. More occupancy, more use, faster degradation.

Moving Forward

Not only do we produce these reports and long term studies, we also build and structure the network framework that collects data, analyses and cleans the data and provides long term asset plans, frameworks to be able to improve your building performance over time.

Get in touch if this sounds like something that is of interest to you or your asset either from an FM provider side or from the asset owner side.

Call Us Now to Discuss a Solution