The noise about the next big thing can make it difficult to identify which technologies truly matter. I will attempt to sort through the many claims to identify the technologies that have the greatest potential to drive substantial economic impact a01e92d4c5acb2beabab043bffd1baf4f7d913bb0cend disruption before or by 2030(our country’s vision set plan). Important technologies can come in any field or emerge from any scientific discipline, but they share four characteristics: high rate of technology change, broad potential scope of impact, large economic value that could be affected, and substantial potential for disruptive economic impact.

Many technologies have the potential to meet these criteria eventually, but our leaders(CS Mr.Mucheru of ICT) needs to focus on technologies with potential impact that is near enough at hand to be meaningfully anticipated and prepared for. I already love the whitebox launch  initiative by his ministry and hope to see such moves in several other ministries and NGOs as well.Therefore, as a country,we need to focus on technologies that we believe will have significant potential to drive economic impact and disruption by 2030.

What Safaricom,Liquid telecoms,Airtel and Telkom have done in enhancing mobile Internet, for example, has affected more than 20 million Kenyans going about their lives, giving them tools to become potential innovators or entrepreneurs— making the mobile Internet one our most impactful technologies. I personally an a beneficiary of their noble initiative to nature and grow talent within STEM.My focus for now is profoundly learn the Internet of Things technology which will connect and embed intelligence in billions of objects and devices all around the world not Kenya alone, affecting the health, safety, and productivity of billions of people.I want to be part of the few folks who will go down the history books of STEM in Africa that shaped the path of the four disciplines and made life of the coming generation better from our engineered solutions.

Here is what I sorted out as the buzz in-things for Tech evangelist, engineering gurus and STEM lovers which will definitely shape our country in the next few years and propel the big 4 agenda to realization.

  1. Mobile Internet Increasingly inexpensive and capable mobile computing devices and Internet connectivity
  2. Automation of knowledge work intelligent software systems that can perform knowledge work tasks involving unstructured commands and subtle judgments
  3. The Internet of Things Networks of low-cost sensors and actuators for data collection, monitoring, decision making, and process optimization
  4. Cloud technology Use of computer hardware and software resources delivered over a network or the Internet, often as a service
  5. Advanced robotics increasingly capable robots with enhanced senses, dexterity, and intelligence used to automate tasks or augment humans
  6. Autonomous and near-autonomous vehicles Vehicles that can navigate and operate with reduced or no human intervention
  7. Next-generation genomics Fast, low-cost gene sequencing, advanced big data analytics, and synthetic biology (“writing” DNA)
  8. Energy storage Devices or systems that store energy for later use, including batteries

9.3D printing Additive manufacturing techniques to create objects by printing layers  of material based on digital models

  1. Advanced materials Materials designed to have superior characteristics (e.g., strength, weight, conductivity) or functionality
  2. Advanced oil and gas exploration and recovery Exploration and recovery techniques that make extraction of unconventional oil and gas economical
  3. Renewable energy Generation of electricity from renewable sources with reduced harmful climate impact

The technologies on my list have great potential to improve the lives of billions of people, starting with our 45 million plus population. Cloud computing and the mobile Internet, for example, could raise productivity and quality in education, health care, and public services. At the same time, some of these technologies could bring unwanted side effects. The benefits of the mobile Internet and cloud computing are accompanied by rising risks of security and privacy breaches. Our 12th parliament ought to sit more frequently and speed up the data protection bill, the computer misuse and cybercrime bill and improve our information and communication act. Objects and machines under the control of computers across the Web (the Internet of Things) can also be hacked, exposingfactories, refineries, supply chains, power plants, and transportation networks to new risks


In considering the disruptive potential of these technologies, I foresee that each could drive profound changes across many dimensions—in the lives of Kenyan citizens, in business, and across the global economy. As noted from key speeches from our president Mr.Kenyatta and his deputy Mr.Ruto, the future seems bright for entrepreneurs and innovators. 3D printing, the mobile Internet, cloud technology, and even next-generation genomics could provide the opportunities and the tools to allow small enterprises to compete on a meaningful scale and advance into new markets rapidly.Almost every technology on my list could change the game for businesses, creating entirely new products and services, as well as shifting pools of value between producers or from producers to consumers. Some, like automation of knowledge work and the mobile Internet, could also change how companies and other organizations structure themselves, bringing new meaning to the anytime/ anywhere work style. With automation of knowledge work tasks, organizations that can augment the powers of skilled workers stand to do well.


As these disruptive technologies continue to evolve and play out, it will be up to business leaders, entrepreneurs, policy makers, and citizens to maximize their opportunities while dealing with the challenges.Lets make our mother country great and change the face of Africa!


Compiled and authored by : Samwel Kariuki

9th August 2018.


While still understanding the changing needs of the market over the years, I feel there is evidence around the globe that providers of energy management produc


ts and services have started taking stock of their offerings in order to include additional reporting and analytical features. Clearly, the aim is to help businesses(most if not all) achieve a wider objective. In fact, some companies like Safaricom but here in our own country have for example seriously started considering an extension of their current capabilities; While that may be one way of effecting better energy management, some providers(mostly international) are heading down the path of integrating their product lines to offer a single solution. A case in point is IBM, which is trying to address the issues raised in a forum, Green Sigma™ Coalition, formed by like-minded companies. This forum aims to help organizations become more efficient and sustainable.

IBM seeks to create intelligent systems that optimize resources, including smart grids, water and traffic, not only at the macro level but to also provide time-relevant data to gain insights and inform forward-looking decisions.


Such new capabilities can reduce the dependency of businesses on external agencies/consultants. Taking such approaches can enable companies to automate, monitor and control the two-way flow of energy from power plant to plug, create transportation systems that optimize traffic flow and decrease CO2 emissions, provide advanced portfolio/task management capabilities, use predictive analytics for better management of resources and suggest corrective action.

Kenyan manufacturers stand to benefit greatly from technologies that various energy-management service providers are implementing to deliver an automated platform (with intelligent alerts and modeling tools) for data collection,

consolidation and analysis. With some of these new methods, the energy management market in our country will advance to provide intelligent analysis once energy usage data is seamlessly integrated with other systems across work streams, such as supply chains, plant controls, financial reporting systems, material planning and product design, waste management, facility management and smart grid/meters. The best way to achieve such advancement in the industry is perhaps consolidation through acquisitions of niche energy management consulting companies by large ERP or IT product companies. Small players offering niche services will eventually realize that the only way to avoid becoming obsolete with the change in paradigm is to be acquired by larger companies since most businesses seek enterprise-wide solutions with a broader set of capabilities than mere energy monitoring. This will help unlock a new environment for energy management that will empower businesses to leverage sustainability investment opportunities within their value chain. They will be able to develop custom dashboards and reports and share relevant metrics with their customers, regulatory authorities and investors — leading to greater accountability. As an outcome, businesses will not only be able to lower costs and increase efficiency, but they will cap wastage and pollution, leaving a cleaner environment and more abundant resources to future generations.


Compiled by: Samwel Kariuki


The Internet of Energy (IoE) can be broadly defined as the upgrading and automating of electricity infrastructures, making energy production more clean and efficient, and putting more power in the hands of the consumer.WhatsApp Image 2018-05-31 at 15.45.30 (1)

My blog today will discusses how to apply ML analytics in the utilities industry to create the IoE.I personally choose to see IoE as one system where data in Kenya will be shared and analyzed, producing targeted, efficient results to utilities and consumers across our country.

The first major utility sector is Generation, which relies heavily on the work of turbines. Turbines, whether they be fueled by natural gas,steam, nuclear, or coal, are massive engineering marvels from a mechanical standpoint. There are thousands of moving parts with extreme tolerances, and minute disturbances in the system can lead to major problems, causing downtime, loss of power, safety concerns, and more.

In our country, many grids are plagued with unreliable service. This is primarily because of aging equipment; poor maintenance; and in many cases, the struggle to upgrade power systems to keep up with very high annual demand growth rates. Investment in IoT for both existing and new equipment has the potential to significantly reduce unscheduled downtime by identifying problems before they occur, thereby improving reliability and reducing costs. Other applications of IoT are optimal use of generation assets to increase the efficiency of production. In conventional power plants, IoT would be used to tune the operation of a power plant in real time and to balance production with life cycle cost of maintenance and life of equipment. As an example, GE is about to launch a digital power plant systems for coal plant in Lamu. GE claims its digital technologies when applied to new coal and gas fired power plants will increase fuel efficiency by 3%, power output by 2%, and reduce unplanned downtime by 5%, operation and maintenance costs by 25%, and fuel consumption during starts by 20%.6 In Kenya, these strategies may be used to reduce cost of electricity production and emissions. Another good example of IoT use for optimization of operations is in the wind power industry where (i) wake losses are reduced in a wind farm by adjusting pitch and yaw angles of individual turbines, (ii) turbines production is increased above rated value in a controlled manner as long as the stress and fatigue loading are within acceptable limit, and (iii) settings of individual turbines are optimized to local conditions to increase output. GE claims a 5% to 10% increase in annual energy production with these strategies.7 A futuristic application of IoT is a holistic optimization of the entire power network with the goal of decentralization and defossilization of the power sector. IoT has the potential to achieve such a transformation in which (i) renewable energy is generated close to load centers; (ii) energy storage devices are used to store excess energy and deliver energy during periods of high demand; (iii) demand response is used to balance supply and demand; (iv) flexible centralized fossil fuel-based power plants plan production based on real-time predictions of variable renewable generators; and (v) dispatch logic, and controllers are used to manage the flow of power. Several of these transformations are being tested in a number of pilots in our beloved country with the goal of achieving close to 100% renewable energy in the power sector and IoT will be a key enabler.


Happy Madaraka holidays!


Complied by: Samwel Kariuki

I think the main challenge while developing the framework to determine the impact of IoT projects on the SDGs in our country is finding ways to compare projects on a “like-for-like” basis. This is difficult to achieve given that projects vary in purpose, geography, size, focus area, etc.


Just the other day i was leaving upperhill for town and along uhuru park i took a sit on one of the benches and started to ponder how is it that we have brilliant sharp folks around with hubs all over yet we haven;t yet implemented even one solution via IoT to assist our fellow Kenyans.

City Digital in Chicago for example has developed a Smart Green Infrastructure Monitoring project which helps reduce urban flooding and prevents property damage, using better informed capital planning for infrastructure investments. Isn’t this what we need for our country right now to save our brothers and sisters in flooding areas???

I wont write more today for i feel for the less fortunate  who cant work out their maths via tech and use it to improve their lifestyle and lively-hood…….but i will play my part now that CAK has done something in line with ICT developments touching on the newest disruptive technologies being implemented around other countries.

My fellow millennial,lets put our energies together and improve our country plus continent via IoT and all other forms of disruptive techs………at least for mankind if not for nature or both.


Complied by: Samwel Kariuki

Early this year there was a symposium titled “AI/IoT-realized Super Smart Society and Energy Network” and was sponsored by the International Research Center of Advanced Energy Systems for Sustainability (AES Center), Institute of Innovative Research, Tokyo Institute of Technology, the symposium dealt with how artificial intelligence, the Internet of Things and other advanced information technologies would transform society and the energy world and what business chances and challenges would emerge, as indicated by the title.image
The symposium consisted of three parts – Part 1 “National Strategy and Outlook on Super
Smart Society,” Part 2 “Super Smart Society and Energy Technologies Seen from Academia,” and Part 3 “Panel Discussion – Social Implementation Led by the Private Sector.”
What is the “super smart society?” This is an interesting question. In Part 1, it was argued that human society historically transitioned from a hunter society to a farmer society, an industrial society and an information society, or from Society 1.0 to Societies 2.0, 3.0 and 4.0, before a new economic society comes as Society 5.0 or super smart society. The new society was explained as a society in which AI, big data, IoT and other advanced technologies would be fully used to achieve both economic development and the resolution of social challenges facing the world. The super smart society was also described as a society in which AI, big data and dramatically advanced information technologies (electronics, communications and data processing) would be fully used to
integrate cyberspace with physical space to produce new values.
An important challenge in energy and environment areas in our country and the whole Continent at large would be to build a low-carbon society and very efficient energy supply systems using renewable energy, storage batteries, hydrogen,advanced next-generation vehicles, distributed energy systems, demand response systems, virtual power plants and other technologies. AI, big data, IoT and other advanced technologies would be fully used to digitize and expand the energy world in the new economic society. As a matter of course, the super smart society and energy’s expected roles in such a society represent long-term strategic challenges, with any specific path to such society remaining uncertain(we have a tendency of assuming things until they turn out to be a necessity in our day to day lives). There may be numerous problems to be resolved for realizing the new society.
Nevertheless, initiatives to overview social transition and transformation from a broader perspective and depict and pursue the future society we should build are very significant and valuable. We will have to closely watch future initiatives to realize the super smart society and energy’s roles in such society. Based on matters of interest to me and my expertise, I strongly believe the Kenyan super smart society would be digitized and electrified, energy security (particularly, power supply security) would be the key to realizing and managing most of the activities. I have noted three points on new risks that we as Kenyan engineers would have to consider in regard to energy and power supply security while digitization and electrification would make irreversible progress.
The first point is the impact that the substantial expansion of renewable energy including intermittent solar photovoltaics and wind power generation would exert on power supply systems.
Storage batteries, grid enhancement measures, auxiliary fossil power generation and other measures are required to cover the intermittency of solar PV and wind power generation. This means additional costs. While solar PV and wind power generation costs are remarkably declining, the additional costs are required for integrating such intermittent renewable energy into power supply systems and may not necessarily be negligible. As intermittent power sources’ share of the power mix expands further, the costs for integrating these sources into power supply systems will grow. Power supply security and the integration costs could be challenges.
The second point is related to cybersecurity since am a member of KCSFA (kenya Cyber Security and Forensic Association) and i follow closely our own internal debates and discussions. As social and economic systems grow more dependent on stable power supply due to further digitization and electrification, they are likely to become more vulnerable to cyber attacks. As cybersecurity problems are growing more complex in our country, diverse and serious, cybersecurity measures must be updated in response to the fast-changing situation. So far, cyber problems have not become as serious as the oil crises that globally shook energy and power supply. As stable power supply becomes the most important challenge in the digitized society, however, we should recognize
cybersecurity problems as a major potential risk. The third point is a stable power supply problem related to investment in deregulated markets.In Kenya, power and gas system reform will need to be implemented to deregulate markets more and more
through the beginning of the 2020s(Lets stop thinking only politics in 2022). In globally known cases, there are many cases where investment costs in power sources has failed to be recovered in deregulated power markets, leading to the so-called “missing money problem”. The classic “missing money problem” has transitioned to a more complex problem as wholesale power market prices have declined due to the large scale inflow of renewable energy power generation promoted by policy support and cost reduction. In response, the introduction of the capacity mechanism is being considered or implemented. In the digitized and electrified society, how to secure investment and stable power supply in liberalized markets with renewable energy expansion trends may be a key challenge.
While great expectations are placed on the realization of the super smart society, or Society 5.0(as i would love to call it),there are many challenges to tackle in the energy world in our continent. In the new economic society in which advanced technologies are fully used, energy is likely to take an even more important position instead of staying at its present level of importance. Energy security will thus remain an old and new issue.
                                                                            Complied and written by : Samwel Kariuki

Innovations in science and technology have allowed our Kenyan telecommunications industry to expand services, reduce prices, and grow to meet the ever increasing demands from other sectors—sectors that increasingly rely on telecommunicatio


ns to help them work efficiently while reducing their energy consumption. The current worldwide focus to reduce energy consumption and ecological impacts is adding another dimension to the business cases for new network technology, implementation of energy-saving systems, and installation of green alternative power solutions. All these requires a plan based on network architectures, service offerings, customer base, and geographic location. This is a challenging, multi-dimensional planning process that I personally am very eager to take up as an engineer and give solutions to help my fellow Kenyans and the entire country at large. Planning an efficient, eco-sustainable telecommunication


s network while meeting corporate financial and business goals is a challenging problem. While networks are naturally evolving more so in east Africa to become more efficient, this am afraid will take years to occur(note as a country, we have made bigger stride


s on the evolution of our networks as per CAK 2017 report). With the current high energy costs and government and consumer interest in greener technology, proactively reducing energy consumption and adding alternative, eco-friendly power sources often result in a better business plan, both short term and long term. Deciding on specific energy reduction initiatives and determining the viability of alternative power options that meet network operator business goals require a complex, multidimensional analysis that I should clearly state forthright that Safaricom PLC has already identified these fields and has started off tapping into giving solutions aligned together with the Paris agreement as well as the 2015 SDGs.



My personal goal: I want and will come up with a solution(s) to help Kenyan network operators’ work through the maze of conflicting requirements and options to identify eco-friendly solutions that meet the financial and business goals while reducing the ecological impacts.


Written and complied by: Samwel Kariuki

It is a worldwide goal to reduce energy consumption and CO2 emissions. The EU has targeted a reduction of 20% for year 2020 and just the other week we saw an MoU between Safaricom and UN signed championing for SDGs set.. A part of this energy reduction scheme concerns the    telecommunication industry and ICT that participates in a direct, indirect and systematic way.  Characteristic  examples which are yet to be in full use or are at nascent stages in our country  are  green networks,   smart   buildings, 20170820_152219.jpg  smart   grids,   Intelligent Transportation Systems (ITS), energy efficient electronics (OLEDS, photonics, nanotechnology) and the application of  embedded  systems  towards  low  carbon  and  energy efficient technologies .

Telecommunication networks constitute a major sector of ICT and they undergo a tremendous growth. Capacity issues and delivery of complex real time services are some of the main concerns that yield high power consumption patterns. In our increasingly competitive mobile telecommunication sector, operators are turning to emerging markets for their next step growth situation that increases the number of subscribers and required base station equipment-case examples include safaricom now on 4G+ while Telkom is rolling our 4G across its country’s network footprint. This creates the need for equipment installation to areas where off grid renewable energy solutions are required and energy efficient networks are important e.g. Northern parts of Kenya. In addition, the increase of fuel and electricity costs bounds the OPEX of the system.

Telecommunication networks and broadband access are proved to consume a huge amount of energy for data delivery.  In general, the telecommunication   sector accounts for approximately 4% of the global electricity consumption (I researched widely from ITU web links).  I personally believe that  reduction  of  CO2   emissions  can  be accomplished  by  focusing  on  innovative telecommunication services like online taxation, video conference,   online   billing   that   can   enable   a   green economy. The goal is to deploy telecommunication networks enabling power efficiency, yielding a small ratio of required Watts per Gbps and Watts per user. Green initiatives have already been commenced by different operators around developed countries.

This summarized word press discusses and proposes various energy efficient techniques for the green operation of telecommunication networks. Cellular networks that suffer most of the power waste nowadays are what I will highlight briefly. It is observed that almost 50% (including the operation of servers) is due to the operation of telecommunication networks. These can be mobile networks, WLANs, LANs and fixed line networks. As  far  as  the  overall network  performance  is  concerned  the  energy consumption is higher at the access part of the network and the operation of data centers that provides computations, storage, applications and data transfer in a network. On the other hand, backbone and aggregation networks present lower energy demands. This makes clear that an energy efficient architecture should focus on intelligent and efficient access techniques and efficient operation and data manipulation by data centers. The main functionalities of a network can be summarized as the process of regeneration, transportation, storage, routing, switching and processing of data. The power consumption patterns of these processes can be observed that the largest part of energy is consumed for routing/switching, regeneration and processing of data. Both communication protocols and electronic devices are responsible for this consumption and this imposes challenges for more sophisticated transport techniques, thermal removal from switches or the servers and less redundant data transfers.

For mobile networks, a crucial factor affecting network power consumption is the site operation that incorporates base station equipments. . It is obvious that the greatest portion of energy is consumed for cooling of equipments and base station operation. Monitor operation and lighting requires the minimum of energy whereas for the backhaul energy consumption the picture is not clear and depends on the type of connections of the backhaul network (fiber or cable).  Within the base stations, high power demands are due to feeders (transmission of radio waves), the RF conversion units and power amplifiers, signal processing units and various electronic   equipments   such   as   air   conditioners   and auxiliary equipments.

The power consumption within a base station exhibits important similarities with data centers. The available power from the electricity grid, the battery backup unit or the renewable energy (RES) enters the base station and is divided into an in-series path and an in-parallel path. Non- critical equipments support the operation of the IT equipments that are divided into radio units and baseband units. The most energy consuming devices of base stations are the cooling infrastructure, power amplifiers, RF feeders and the AC/DC and DC/DC conversion units. Depending on the number of sectors, nSC, and the antenna number, nTX, of the base station, the total power consumption is computed as follows;



In the above formula an additional factor models the power consumption due to RF links of the base station. For macrocell and microcell base stations, empirical formula can describe the relationship between the power delivered to the antenna relative to the consumed power of the base station [13]. For macrocell stations the power consumption is almost independent of the input load (traffic) whereas for microcells, power consumption is highly dependent on the input load.

Making a network to operate in a green manner is a complex task. Sometimes, optimizing energy consumption in one part of the network can increase power consumption and degrade the performance of another part of the network. In general, total network optimization is better than the sum of optimizations of individual parts. A network to work in an energy efficient way is not only a matter of environmental protection through signing of memorandums but also a crucial factor for the deployment of future networks to off grid areas that rely on Renewable Energy Sources (RES) or personal and sensor networks that rely on battery power supply. Minimizing power consumption has also a great effect on the cost of operation of a network and this makes it more affordable to the user. Network energy efficiency can be considered as a very complex task since there is no clear solution to the problem. The sectors of the network that require the greatest attention are the electronic equipments of both end user and the access network, thermal removal processes, efficient network planning and base station design.


Compiled and written by: Samwel Kariuki

Technology is changing rapidly for wireless, significantly changing the power requirements of the 6000+ base stations within our Kenyan Telcos infrastructure. These improvements increase the viability of using Eco-friendly power and our Telcos have img-20161025-wa0004.jpgalready seen this trend of IoT and are engaged in efforts to stop the trend of rising telecom energy demands. With so many options for reducing their Eco-footprint, and considering the challenge of implementing changes while remaining profitable, planning a sensible, ecologically friendly path forward is often a formidable task. It is for this reason that I chose to take an opportunity to write to the power departments in our communication institutions which I have gracefully worked with for close to 3 years indirectly as an engineer assigned to do electrical and computational works for them.

The4G+ as an already laid out plan by one of the major Telcos within our country serves as an example which is a really good move that comes with growth of bandwidth demand which can easily cause Safaricom network  energy  consumption to  rise  in  step  with  the growth. The resulting increase in electricity costs leads to reduced margins at a time when competition is also driving prices down-the relauch of Telkom Kenya a few days ago marks a threat in the same regard. Having worked closely with a number of power departments amongst the Telcos we have, I have seen and learnt two options used when planning to reduce power consumption.  First, there are new network architectures that are inherently more energy-efficient and which can simultaneously provide the flexibility to support continued increases in demand. Second, choices in network equipment, options, and support equipment for new or existing infrastructure have also had a tremendous impact on the amount of power consumed. Both options are quite viable and should be part of any power reduction plan even as we leap into the digital disruptive era in the coming years.

Am grateful to have worked indirectly with the engineers at both power and optimization departments and have been able to tap a lot of skills in my area of expertise and personal growth as well. I look forward for an opportunity to present my ideas (a combo mixture of Artificial intelligence, big data analytics and IoT) as well as deliberate further on how best can power can be planned and supported to attain the ultimate goal in energy efficiency. Am also grateful to Parastatals that deal directly and indirectly with power and energy distribution for the nifty work they are putting across to solve the trilemma of cost, reliability and quality of power being used in our republic.

Below is a recap article of the latest bell lab power technical journal 2017 edition that I saw it prudent to share as well with other engineers and stakeholders in power & energy sector alike whom I revere and hold atmost respect for the trainings and lessons I have gained from them.


Methodology for Planning Energy-Reducing

The methodology for planning network changes to reduce energy usage consists of three cascading steps:

  • Energy consumption hierarchy. Identification of the network elements that consume power and their location in the network.
  • Energy-saving chain. Identification of network element dependencies upon each other’s  power  dissipation (e.g.,  larger  air  conditioning units  having  higher energy   consumption are  necessary if inefficient power   rectifiers   are   installed  because  of  the energy  they  waste  through heat  radiation).  This allows network operators to target the most effective points for energy reduction by applying energy-saving initiatives.
  • Energy-saving initiatives or options. Determination of specific choices or actions that  can be taken to reduce energy  consumption for one or more  net- work elements (e.g., replacing low-efficiency rectifiers with high-efficiency rectifiers, which requires capital  and  installation expense,  but these  expenses may be offset in 12 to 15 months based  on  today’s  high  energy  costs).  Sets of initiatives are often deployed simultaneously due to typically lower installation costs as compared to deploying the initiatives one at a time


As we continue improving our communication systems across the country and beyond, lets research and read widely for the upcoming 4th industrial revolution which in my own view will be sparked and born in China & embraced fully here in Africa and hopefully in our dear motherland Kenya.




“A powered nation is a growing nation”~Samwel Kariuki

Data Mining is the process of sifting through stores of data to extract previously unknown, valid patterns and relationships that provide useful information. Once these patterns are found they can further be used to make certain decisions for development of businesses. For decades major components of data mining img_2337.pngtechnology have been under development in research areas such as statistics, artificial intelligence and machine learning. But in recent times we are yet to see here in Kenya the maturity of these techniques coupled with high performance relational database engines and broad data integration efforts make these technologies more effective for current data warehouse environments within our telcos.

Telecommunication data pose several interesting issues for data mining more so in our now-growing digitalization exodus that a lot of Kenyans if not most are experiencing. The first concerns scale, since telecommunication databases may contain billions of records and are amongst the largest in the world. A second issue is that the raw data is often not suitable for data mining. For example, both call detail and network data are time-series data that represent individual events. Before this data can be effectively mined, useful “summary” features must be identified and then the data must be summarized using these features. Because many data mining applications our telecommunications industry involve predicting very rare events, such as the failure of a network element or an instance of telephone fraud, rarity is another issue that must be dealt with. The fourth and final data mining issue concerns real-time performance: many data mining applications, such as fraud detection, require that any learned model/rules be applied in real-time.

We can distinguish three main steps of describing data mining problems:

  1. Problem formulation in the domain terms. This is usually textual description of the business requirements that have to be fulfilled by data mining.
  2. The transformation of business requirements into a class of data mining problems like classification, prediction, associations etc. It is a bridge between business description and detailed model specification.
  3. The detailed model specification. This is a model specification that is used by data mining modeler for a specific software tools.

It can be observed that one of the main areas of applications of data mining on business level is a support for various tasks of the marketing departments. The data mining becomes a key part of analytical subsystem of customer relationship management systems. On business level of analysis there are many similarities to other industries. The following main problems for example in marketing and sales departments of telecommunication operators can be distinguished: – customer segmentation and profiling, – churn prediction, – cross selling and up-selling, – live-time value, – fraud detection, – identifying the trends in customer behavior. On product or service level there is a focus on analysis of incomes, quality of the service, grade of the service and others. There are formal agreements called service level agreements (SLA) between providers of the service and the customers. Service level management (SLM) is becoming the prevailing business model for delivering a products and services. Such approaches need advanced computerized tools. On the level of infrastructure and network analysis we can distinguish the following problems: – network planning, – IT resources planning, –: fraud detection, marketing/customer profiling and network fault isolation.

One central issue is that telecommunication data is often not in a form—or at a level—suitable for data mining. Other data mining issues that are or/and will affect our Kenyan telcos include the large scale of telecommunication data sets, the need to identify very rare events (e.g., fraud and equipment failures) and the need to operate in realtime (e.g., fraud detection). Data mining applications must always consider privacy issues. This is especially true in the telecommunications industry, since telecommunication companies maintain highly private information, such as whom each customer calls. Personally am elated that most telecommunication companies in our country utilize this information conscientiously and consequently privacy concerns have thus far been minimized. A more significant issue in the telecommunications industry relates to specific legal restrictions (provided by CAK) on how data may be used. In the United States, the information that a telecommunications company acquires about their subscribers is referred to as Customer Proprietary Network Information (CPNI) and there are specific restrictions on how this data may be used. It’s in Kenyans interest that the communication Authority of Kenya (CAK) generally sticks to prohibiting the use of that information without customer permission, even for the purpose of marketing the customers other services. In the case of customers who switch to other service providers, the original service provider is prohibited from using the information to try to get the customer back (e.g., by only targeting profitable customers). Furthermore, all our CSP (communication service providers) & ISP (internet service providers) should be prohibited from using data from one type of service (e.g., wireless) in order to sell another service (e.g.. landline services). Thus, the use of data mining is restricted in that there are many instances in which useful knowledge extracted by the data mining process cannot be legally exploited. Much of the rationale for these prohibitions relates to competition. For example, if one of our telcos can leverage the data associated with one service to sell another service, then a number of Kenyan companies (both CSPs & ISPs) that provide fewer services would be at a competitive disadvantage.

Kazi kwenyu CAK as we anticipate for a colossal unprecedented digital migration that will in hand come bearing babies like data mining, Artificial intelligence, cognitive cloud computing and all sorts of intelligent smart technologies with them. As for our Kenyan telcos, I strongly believe having worked in this industry for more than 3+ years that platforms and foundations (software and hardware) are already set in place in anticipation for the coming 21st exodus.


Complied by: Samwel Kariuki

Dated: 23rd May 2017

There has been recent progress in the analysis of call-center data.   Call-by-call  data  from a small number  of sites  have  been  obtained  and  analyzed,  and  these  limited  results  have  pr…

Source: Future Work in Data Analysis and Forecasting within our Kenyan Telcos and Africa at large.