BearingPoint as the partner of the ESCP “Retailing 4.0" Chair is committed to contributing to building bridges between the academic world and companies. As part of this engagement, the BearingPoint Retail, Luxury and CPG team mentored the writing of a series of articles of ESCP MSc students and shared insights on major retail stakes. Enjoy your reading !
IoT (Internet of Things) refers to any devices or products that can be connected to the internet (other than the usual Internet connected devices like computers and smartphones). These devices can be monitored or controlled from a distance (short and/or long distance) and are able to collect and share electronic information using Internet Protocol (IP).
The main aim of IoT is to have devices that self-report instantaneously, increase productivity and improve information gathering.
In the retail world, IoT has a big role to play, those technologies can enhance performance, reduce costs, increase growth while improving the customer experience.
In the case of IoT, the most important feature that can be considered is connectivity. Without seamless communication among the interrelated components of the IoT ecosystem (i.e. sensors, computers, data hubs, etc.) it is not possible to execute any proper business use case. Not only devices can be connected, but also objects and products can. IoT devices can be connected over radio waves, Bluetooth, Wi-Fi, Li-Fi, etc. Various protocols of Internet connectivity layers can be leveraged in order to maximize efficiency and establish generic connectivity across IoT ecosystems and industries (Educba, n.d.).
A frequently used technology is Radio Frequency Identification (RFID). This is a technology where tags are attached to objects. These tags can contain more information than traditional barcodes (Lee & Lee , 2015) and make it possible to identify them since they own a personal identification and memory which can contain product numbers for instance. The RFID reader generates electromagnetic fields which enables the location of tagged products within a range and exchange data (Ajami et al., 2013). The data is stored in the tag under Electronic Product Code (EPC), which is an identification system which is globally applied (Lee & Lee, 2015). RFID enables multiple applications from stock information e.g. availabilities, quantities, location, etc. to when reorders must be placed (Ajami et al., 2013). Thus, improving operational efficiency and inventory management after implementation. To illustrate, one of the main advantages of this technology is to perform bulk RFID tag reading instead of reading items one by one using 2D code technologies (QR, barcodes, data matrix).
There are three different types of RFID tags.
As humans can collect information about their surroundings thanks to their 5 senses sending signals to our brains, so does IoT. Sensors are implemented to collect signals which are converted into data in such a way the system can drive meaningful insights and appropriate business decisions can be taken. Sensors are able to capture all kinds of physical signals in order to measure for instance light, temperatures, tires pressure, etc. and convert them into data. To make a case successful, it is required to choose the proper combination of sensors and keep in mind the measures and data will vary continuously over time. (Educba, n.d.)
For instance, in cases where temperature must be controlled, the data will change over time according to the weather conditions and locations. IoT devices should be designed taking into consideration these variations and have to be resistant to extremely low or high temperatures.
In 2018, Karim et al. did an experiment about monitoring food storage humidity and temperature. This illustrated how automated measuring of these parameters can lead to evaluation and monitoring of this data on a mobile phone or PC. When a breach in the cold chain takes place, automated turn on/off of an external switch can condition the temperature and humidity in the cold storage. This can prevent food items from perishing when not being stored in the right temperature, e.g. meat.
As defined by Lee & Lee (2015), cloud computing makes it possible to have on-demand access to a shared platform of configurable resources such as computers, servers, storages or software. Since an IoT system is generating a huge amount of data, important data storage capacities are required. Also, a fast broadband is needed to stream the data. Cloud computing is an answer to those two requirements (Lee & Lee, 2015).
As scaling is key, the sensors and the devices are collecting and generating a large amount of data that need to be stored. Current data storing systems are not able to collect and process this personal and heterogeneous data. Companies need to invest in storage facilities to collect all the generated data (Lee & Lee, 2015). Moreover, technically advanced data analytic tools will have to be implemented, be able to support various type of data and translate them to meaningful insights.
From end components to connectivity and analytic layers, the whole IoT ecosystem demands a lot of energy. While designing an IoT ecosystem, the energy consumption level should be maintained as low as possible. Also, the collected data must be stored in data centers and this can be energy consuming. As mentioned before, some sensors are using batteries (Lee & Lee, 2015), but meeting the energy requirements for the whole IoT system remains challenging (Georgiou et al., 2018).
Lastly, data protection and safety are one of the main features of the IoT ecosystem. Indeed, sensitive information like purchase or location is flowing in the network thus designing an IoT system with proper safety, security measures and firewalls is required to keep the data away from misuse and manipulation. Also, by adding more and more devices and products to the IoT system, the security threats increase (Lee &Lee, 2015).
Currently, in almost every IoT use cases, the data is used to make key business insights and drive important business decisions. The generated data can for instance give insights about a changing consumer behavior. There is a need for software and engineers to analyze the data appropriately (Lee & Lee, 2015). As per Google, only 50 % of structured and 1 % of unstructured data is used to make business decisions. While designing the IoT ecosystems it is essential to consider the future needs of analyzing important data quantities to meet incremental business requirements (Educba, n.d.).