Technologies for Shared Office Space

An average of 50% of the available space in a typical office building remains unoccupied over the course of a business day. At the same time, space in cities is getting very short and incredibly expensive. To counter this trend, services for shared office space are becoming more and more popular. Beside flexible and adaptable room layouts, there is a need for data infrastructures that enable demand-based intelligent control of all building areas. Thereby, not all technologies are equally suitable for solving the requirements of such an agile concept.

Co-working has had a major impact on the performance and utilization of commercial real estate. It has been responsible for a significant amount of positive net absorption in major metropolitan markets enabled by additional investments into upgrades and expansions of amenities in existing buildings. From humble beginnings in 2007 with just 14 of such spaces in operation, it was estimated that during 2018 more than 600,000 people would be working in 4,500 co-working spaces in the US alone. The worldwide number of such spaces was expected to top 15,000 by the end of 2018 [1].

Several key drivers have enabled this development – easy access to office space within the desired geography, the wish for flexible working environments that can be easily adapted to the changing number of employees and the desire to co-work within an open, creative atmosphere that stimulates exchange with other people. This has been accelerated by a trend towards larger co-working spaces (with over 100,000 square feet of area*) that are targeting established companies. By offering office space in target locations with technological resources and talent, they can stay in tune with developing industry trends and changing demands in resources.   

Addressing the requirements of such a diverse customer base is possible only with highly flexible architectures enabling different usage scenarios and easy reconfiguration. Workspaces need to be resizable and reconfigurable (open, semi-open, fully enclosed) while open collective spaces invite social and professional interaction between users, enabling occasional, recurrent, and daily meetings as well as special events.

In short – agile working models require agile buildings.

Agile Buildings Tasks

Agile buildings enabling shared office spaces have to provide a number of key features – an architecture and technical area (building, lighting, HVAC, controls) suited for user convenience with the option for constant reconfiguration, an IT and data infrastructure enabling data analytics to monitor and optimize key performance figures and a service offering (administration, access control, supplies, cleaning etc.) that meets the needs of all involved in a financially efficient way.

User convenience is defined by a wide range of factors including the neighborhood in which the building is located, the internal layout of work and open spaces and the interior design. All of this combined shall create an inviting atmosphere for work and collaboration. Specifically for controls, the individual adjustment of workplace lighting according to user preference is one of the most relevant requirements. Some buildings might additionally allow the control of window blinds and even heating and ventilation by the user, but this is generally an exception.

Constant reconfiguration means that installation cables and equipment into fixed positions should be minimized and as much infrastructure as possible should be placed into the ceiling to minimize obstruction of the floor layout. Data analytics require a flexible sensor infrastructure capable of monitoring key parameters such as desk/space utilization, supply level tracking (e.g. soap and towel dispensers) and demand-based cleaning in consideration of activity and utilization.

Finally, the results generated by these data analytics have to be easily usable within service models both of the building operator (e.g. space utilization monitoring) and external service companies (e.g. cleaning and catering).

New Service Models

Agile environments with changing requirements benefit from service models where non-core activities such as cleaning and catering are outsourced. Only a minority of companies today still employ their own staff for such tasks. Similar trends start to become visible in the building infrastructure as well.

Taking the case of lighting, the benefits of the transition from conventional to LED lighting have enabled new business models where third parties become partly or fully responsible for the operation of the lighting system. Energy service companies (ESCO) that traditionally have focused on the execution of energy saving projects such as lighting upgrades are increasingly offering financing methods where a part of the energy cost saving for the customer is used to pay the project cost.

Under this model, an ESCO will perform an initial analysis of the property and its energy consumption and then design, install and maintain a more energy efficient solution. The saved energy cost will be used to pay back the capital investment of the project over an agreed period of time ranging from five to twenty years. If the return on investment is insufficient then the ESCO is usually required to pay the difference.

Taking this model further, Lighting as a Service (LaaS) is a service delivery model in which light is delivered to the customer according to his requirements and charged on a subscription basis rather than via a one-time payment.

What both models have in common is that the provider is responsible not only for the initial installation of the lighting system but also for its operation. Operational expenses thereby become a critical factor for the return of investment (ROI). The goal of optimizing operation expenses drives a strong demand for data insight allowing for automatic control of building parameters such as light levels based on occupancy and available light or to schedule tasks such as cleaning based on actual utilization rather than fixed schedules. It also requires minimizing maintenance work which will be an important aspect when discussing wireless solutions.

Data Infrastructures for Agile Buildings

Data is the basis of agile buildings. It comes in many forms – it can represent user requests (e.g. setting a light level), allow information exchange (e.g. Internet connection) and monitor key parameters (e.g. occupancy).

Exchanging data requires a suitable infrastructure meeting specific requirements regarding the amount of data (speed, volume) to be exchanged, the distance between the partners exchanging the data, the transport medium (wire, fiber, air) and the available power. In an ideal scenario, one infrastructure would cover all those requirements. This, however, is usually not possible. A look at different available technologies explains why.

Pros and cons of different available technologies:
•    Fiber, Ethernet or copper cables are ideal mediums to transport large
     amounts of data throughout a whole building with minimum latency and
     highest reliability. Their flexibility, however, is very limited as they require
     the installation of dedicated cabling
•    WiFi on the other hand is an ideal choice for exchanging large amounts
     of data over limited distances with greater flexibility. However, its power
     consumption makes it not well-suited for wireless controls and sensors
•    Dedicated low power wireless protocols such as Zigbee, Bluetooth® or
     EnOcean, on the other hand, combine very low power consumption with
     wireless flexibility. They can, however, not be used to transport large
     amounts of data or cover entire buildings

Data infrastructures within a building are therefore, by nature, usually hybrids which can be classified by their transport medium (cable versus wireless) and their use of IP protocol as shown in figure 1.

Wired Versus Wireless Data Transport

Wired solutions can combine high communication reliability with high data transport capacity over larger distances making them an ideal choice as data backbone within a building. Ethernet (standardized within the IEEE 802.3 family) is increasingly becoming the standard choice for this application. Historically, wired connections have been used for all applications even without the need for large data rates. DALI – using a data rate of just 1,200 bit per second – is a typical example of that.

The key disadvantage of a wired data infrastructure is its inherent inflexibility. Its need for dedicated wiring makes quick reconfiguration difficult. Simple tasks such as moving a light switch to a different location requires significant construction work. This is not an option for agile buildings. Therefore, wireless solutions have become increasingly attractive as a complement to the wired infrastructure.
One key challenge when moving to truly wireless solutions (without power cables) is the question of how to provide power to these devices. Batteries have been widely used and continue to be an attractive proposition for cases where initial cost is the prime concern and maintenance for battery exchange is no concern.
Maintenance – A Question of Power Source

The impact of maintenance becomes quickly obvious in larger buildings where thousands of sensors are distributed over several floors and offices. Often these devices are mounted unobtrusively in places that are difficult to reach, e.g. on or above drop ceilings. In these cases, battery exchange is a challenging and time consuming effort.

Energy harvesting technology has meanwhile established a reliable alternative to batteries as an energy source for wireless devices. Self-powered sensors gain all energy needed for their operation from the surrounding environment. The most obvious example is the energy harvesting wireless switch that generates its energy from the kinetic movement of being pressed. Other sources used for self-powered sensors are light (indoor and outdoor) or temperature differences to detect occupancy, light intensity, temperature, humidity, access or even acceleration; to name a few.  

Energy harvesting devices, by their nature, do not require maintenance and can therefore have a positive impact on total cost and operation effort. It also gives installations with energy saving purposes an eco-friendly character by avoiding tons of battery waste. Energy harvesting technology can today support a variety of different low power wireless protocols such as Bluetooth® Low Energy (BLE), Zigbee PRO Green Power and EnOcean sub 1 GHz radio that can complement the IP infrastructure of a building.

IP Versus the Rest of the World

When looking at protocols, an infrastructure communicating from end to end, purely based on IP (Internet Protocol), offers significant advantages. Almost every commercial building today needs an IP infrastructure (Ethernet, WiFi or both). Extending this towards sensors and controls would eliminate the need for dedicated gateways translating between different protocols. IP-based lighting control could thereby directly integrate into the IP infrastructure and become much easier to monitor and administrate.

Power over Ethernet

Power over Ethernet (PoE) is an example of a wired implementation of an all IP infrastructure. The key function of PoE is the elimination of additional cabling whenever a PoE-connected device can be supplied directly via the Ethernet cable. Products such as conference phones and IP security cameras are ideal examples of solutions well-suited for PoE.

More recently, PoE has also been proposed for the lighting infrastructure in buildings. The key challenge there is that Ethernet cabling is optimized to transport vast amounts of data and limited amounts of power. Ethernet cabling is capable of sustaining data rates that can exceed 1 Gigabit per second while lighting control requires comparatively little data – the well-known DALI protocol, for instance, operates at just 1,200 Bit per second.

Power delivery, on the other hand, is limited in PoE systems by the resistance of the wires within the Ethernet cable. This resistance generates heat and results in voltage drops proportional to the length of the cable. High quality cabling and special Ethernet switches have to be used to maximize power delivery within lighting systems.

PoE also does not meet the mobility requirement of sensors and switches within agile buildings. Having to connect such devices via Ethernet cable makes quick space reconfiguration very difficult. IP infrastructures are therefore usually augmented with other (typically low power) wireless solutions to provide connectivity to devices within the office environment.

WiFi

WiFi would seem to be the obvious answer for that; however due to the protocol overhead combined with the high power requirements it is not well-suited for this task. The header structure of an IP packet (IPv6) illustrates the point (Figure 2). Here, source and destination address alone add 32 byte of overhead.

Figure 2: The example for an IPv6 data packet demonstrates the data overhead from the source and destination address

Additional requirements such as certificate-based security schemes will further increase that discrepancy. Most IP-based wireless protocols therefore aim to reduce the overhead of IPv6 protocol in the wireless communication.

6lowPAN

6lowPAN has become a common choice for such protocols and has gained a lot of attention recently due to its integration within Thread. 6lowPAN provides IP communication on top of IEEE 802.15.4 (which is also used in Zigbee) as shown in the following illustration (Figure 3).

Figure 3: 6lowPAN integration structure within Thread

This approach combines the benefits of an IP-based communication model with those of a protocol optimized for wireless communication. One immediate disadvantage is that dedicated gateways between Ethernet-based infrastructure and 6lowPAN/Thread are required.

EnOcean

Considering the fact that wireless protocols will need gateways anyway, another approach has been to cover larger distances with minimal energy to enable the longest possible operation time for sensors and controls. The most prominent example of such wireless protocols dedicated to building automation and lighting control is the international EnOcean standard (ISO/IEC 14543-3-1X), which is supported worldwide by more than 400 member companies.
While many protocols such as Bluetooth®, WiFi and Zigbee operate in the 2.4 GHz frequency band, this protocol uses the sub 1 GHz radio bands. This provides the advantage of much higher communication distances for a given transmission power due to lower signal attenuation.

This attenuation (path loss) increases for a given distance directly with the frequency according to the following formula:

The use of lower frequencies therefore provides much better coverage for a given transmission power and is thus optimal for protocols targeting lowest energy consumption. Combining this with highly compressed data frames, as shown in the graph (Figure 4), makes it possible to create fully self-powered (energy harvesting) devices.

Figure 4: Data string structure of the EnOcean wireless protocol Bluetooth®

Another common requirement is the ability to directly interact with the wireless devices using smartphones and tablets. Almost all such devices include Bluetooth® functionality, which has become an attractive option, especially with its Bluetooth® Low Energy (BLE or Bluetooth® Smart) extension.

The history of the Bluetooth® radio standard dates back to 1994 when a number of companies looked for a mechanism to replace wired connections. In that year, Ericsson proposed the concept of a wireless connection that could replace the common RS-232 cables used for communication between different devices.

In parallel to that, companies such as Nokia, Intel, IBM and Toshiba were also investigating mechanisms to wirelessly connect devices such as cellphones and computers. These five companies formed a special interest group (SIG) which was officially established in 1998.

In June 2007, Bluetooth® SIG acquired the Wibree Alliance, a Nokia-led initiative that had developed an ultra-low power (ULP) form of wireless connectivity – using much less power than the existing Bluetooth® wireless technology – that could be used for communication between cellphones and accessories. Beginning with Bluetooth® version 4 (released in 2009) this feature – now called Bluetooth® Low Energy – has become part of the standard Bluetooth® stack.

The recent introduction of a mesh network topology where messages can be relayed from the source to the destination via intermediary devices has greatly extended the communication range of Bluetooth® systems allowing coverage of larger areas with a Bluetooth® Mesh network (Figure 5).

Figure 5: Schematic for a Bluetooth® Mesh network structure

The combination of classic Bluetooth® for communication with smartphones and audio accessories, Bluetooth® Low Energy for communication with sensors and other low power devices and a mesh network topology for covering larger areas makes Bluetooth® protocol a very compelling choice as low power data network. Deploying Bluetooth® mesh within the lighting system therefore provides the infrastructure for sensor data within buildings.

Minimizing Infrastructure Upgrades

The need for upgrading the building network infrastructure to support new data-driven services provides a significant barrier. In many environments such upgrades will create detrimental business disruption which might be difficult to justify. Different schemes have therefore been proposed to minimize such disruption by reusing the existing infrastructure. One promising approach is the upgrade of WiFi access points to be able to capture data from low power wireless devices.

Aruba Networks (an HP Enterprise company) is one of the leading providers of network infrastructure such as wireless access points. Aruba ships millions of indoor and outdoor Wi-Fi access points every year for smart home, retail, healthcare, hospitality, education, service provider, enterprise, industrial, manufacturing, airline, and government customers. With their latest software release, Aruba access points can use their built-in BLE radio to receive sensor data from EnOcean's energy harvesting Easyfit Bluetooth® sensors and switches and forward those into the enterprise network.

The ease with which self-powered BLE devices can be added to an Aruba
network - without the need for new IT hardware or costly installation work -
makes the solution very compelling for customers.

Paradigms of Tomorrow

Future agile buildings enabling highly flexible co-working and co-utilization space will require large amounts of sensor data to continuously analyze and optimize usage and operation. This data forms the basis for new service models that will augment or replace existing building operation paradigms.

Creating, capturing and delivering such data requires a flexible network infrastructure that can continuously adapt to new usage scenarios. In most cases such infrastructure will be a mix between IP backbone systems (Ethernet + WiFi) and dedicated low power wireless protocols.

The upgrade of lighting systems towards larger area wireless mesh networks creates a unique opportunity for a high density, low power data network encompassing the entire building. As an intermediate step, reuse of existing IT infrastructure such as wireless access points can enable quick initial deployment without the need for major upgrades.

Operation and maintenance cost continues to be a key factor for the total return on investment. Under many service models this cost has to be covered by the operator thus driving a strong need for minimizing them. Energy harvesting wireless solutions are ideally suited for such cases. They are therefore a key ingredient to providing the flexible building-level intelligence without maintenance for the agile building of tomorrow.

References:
[1]    gcuc and Emergent Research: "Number of U.S. and Global Coworking
        Spaces and Members 2017 – 2022 (December, 2017 Forecast)"
        http://gcuc.co/wp-content/uploads/2017/12/GCUC-Global-Coworking-Stats-2017-2022.pdf