Machine Learning Skills (2)

Part
01
of four
Part
01

Healthcare - Technology Trends

Telehealth and wearables are two healthcare industry technology trends that have gained considerable momentum recently. Telehealth allows the optimization of scarce resources and makes healthcare more accessible. The potential for use in combination with machine learning is unlimited. Wearable healthcare has gained traction as a trend due to the consumer's desire to have more control over their health outcomes. When combined with AI, there is no doubt; lives will be saved.

Telehealth

  • Telehealth has been described as the "epitome in healthcare trends. Telehealth sees medical appointments carried out via video link. Initially, the appointments were primarily doctor orientated, but as the trend has evolved, the technology has been applied to other healthcare providers.
  • One of the advantages of telehealth is it allows greater coverage, with those in rural communities able to access similar services to those in urban environments. In the future, global telehealth appointments will become more commonplace.

Reason it is a Trend

  • Telehealth offers the opportunity for multiple providers in different locations to collaborate, in a way that is not possible if physical appointments are required, to provide an optimal treatment plan for the patient. It ensures the patient receives the best care from the people best qualified concerning the individual's health.
  • Telehealth creates greater doctor efficiency, which is fundamental given the global shortage of doctors. It also ensures that resources are more appropriately allocated and those in rural communities do not miss out.

Machine Learning Skills to Support Trend

  • The relationship between machine learning and telehealth is still in its infancy, but, as it has developed, one thing that has remained constant is that decision-making must stay in the hands of the doctors. Notwithstanding this, AI offers value to the telehealth concept in a way that will optimize doctor involvement, while insuring the patient receives the best possible care.
  • Technology has already been developed that sees AI technology used to gather and collate the patient history and investigations in the first instance, so all the preliminary work has been completed by AI at the time of the telehealth appointment. This ensures that all the information required for the doctor to make a decision is readily available before the telehealth appointment. It minimizes the need to reschedule or delay appoints, and ensures that the information is consistent for each patient, which will improve the quality of care, consistency, and ensure all patients receive the same standard of care.
  • This technology also provides an invaluable resource in the decision-making process due to AIs ability to provide treatment options based on the outcomes of any other number of patients with the same diagnosis.

Deficits that Hinder the Trend

  • Despite a huge volume of information suggesting that AI outperforms doctors, there is still a reluctance to trust machine learning in the healthcare sector. Patients are also slow to warm to the telehealth concept itself, many struggling to overcome the physical distance between doctor and patient.
  • The medical profession presents the biggest issue in the future development of this tend, with resistance levels high among some professionals, who are unwilling to place trust in a machine, insisting that the benefit of a human element should not be underestimated. There is also a school of thought within the medical profession opposed to telehealth, arguing it is essential the doctor examine the patient.

Wearables

  • Wearable health products are not only a top health trend, but they are the top fitness trend currently. Wearable healthcare incorporating smart watches and fit bits has established itself in the global psyche, and the trend toward the adoption of these products shows no signs of slowing.

Reason it is a Trend

  • The trend is driven by an increasing demand from consumers to have control over their own health destiny. In general, there has been a trend in healthcare, over recent years, toward greater patient autonomy. This trend is a manifestation of that greater desire.

Machine Learning Skills to Support Trend

  • Machine learning is, without a doubt, supporting the current trend toward wearables. While the information collected by the wearable provides valuable insights into consumer healthcare, the value that is added when machine learning is applied to this data amplifies the benefit to the consumer.
  • AI can collate and evaluate the data that has been collected through the wearable and apply it to similar recordings. As technology develops, this is likely to incorporate individual demographics, and family and personal medical history. AI will be able to use this information to make health recommendations to the individual user, which maximizes the potential of the technology, in a way that is not possible were there only human input. The time frame for analyzing the data and the cost savings are also considerable.
  • The daily monitoring of vital signs by AI will enable any serious concerns or anomalies to be identified early and, in the long term, save lives.

Deficits that Hinder the Trend

  • As with telehealth, consumer trust remains an issue, however given wearables increasing popularity and the fact consumers are making the choice to use them, it is likely the same levels of resistance will not be seen as the role AI has to play increases.

Research Strategy

To determine the current trends in the healthcare industry, we reviewed a range of scholarly research, industry and market publications, and industry experts. By doing this, we were able to identify the two aforementioned trends. We consider something to be a trend if it has generated multiple articles, multiple industry experts are discussing it, and companies within the industry are adopting it.

Part
02
of four
Part
02

Aerospace - Technology Trends

Given the publicity that the aerospace industry has received regarding its carbon footprint due to emissions, it is not surprising that there is an increasing trend toward the development of zero-emission or electric aircraft. Flexible-wing technology offers benefits that long term have the ability to change the competitive landscape and increase profitability for aircraft owners. Machine learning has been heavily involved in this process through modeling, the analysis of results, and the computation of optimal conditions. While there are still barriers to commercialization of each trend, the benefits each offers means that there is a sense of inevitability to their continued development.

Zero-Emission or Electric Aircraft

  • Zero-emission or electric aircraft are a relatively new trend that is slowly building speed. Electric aircraft do not rely on the burning of fuel for propulsion; instead, they look to an electric motor connected to a number of battery cells.
  • A significant amount of money is being invested into the development of electric aircraft with Boeing, Rolls Royce, and Airbus, just some of the companies currently vying to become the first to operate a commercial passenger flight.

Reason it is a Trend

  • Given the current flight shaming movement, the zero-emission aircraft can not be developed quick enough for the aerospace industry, as the industry faces ongoing criticism regarding the carbon emissions the industry generates. The carbon footprint that the aerospace industry generates is the primary reason for ongoing research, development, and adoption of this trend.
  • Zero-emission aircraft present as a viable alternative to conventional aircraft in light of the sustained public criticism that the aerospace industry is not adequately addressing this issue.

Machine Learning Skills to Support Trend

  • Machine learning is currently used by the aerospace industry to analyze data from thousands of flights taking into account aircraft type, weight, weather conditions, routes, and altitudes to determine fuel efficiencies and requirements. This machine learning is being developed further and is being used by the developers of zero-emisson aircraft to model flight times and distances.
  • The development of lighter, safer, and more energy efficient batteries is key to the economic viability of the electric aircraft market. Machine learning is being used in research in this area. Part of the research process involves experimenting with different chemicals and combinations. AI is used to assist in this process. Data from previous experiments is applied to an algorithm to select the combinations of chemicals that are most likely to produce maximum efficiency.

Deficits that Hinder the Trend

  • Battery technology is one of the main reasons that the development of electric aircraft is slow. Ultimately, battery efficiency determines the flight distance, and current battery technology is insufficient. Progress in this area needs to be made if this trend is to continue to gain momentum.
  • The electric motors currently available are also limiting electric aircraft development and deployment. While there are some issues around the power output of the motors, the larger problem is thermal management, given the heat generated by the electric motor, and the need to develop an appropriate heat extraction system to prevent it damaging the motor itself.

Flexible Wing Technology

  • Question marks exist over the long-term future of fixed-wing aircraft, given the current trend toward flexible-wing technology. A number of tests are currently underway that offer shape adaptability throughout the flight due to the advantages it offers. By adapting the wing at various stages of the flight, the benefits of aerodynamics can be maximized, which has flow-on effects on aircraft efficiency.

Reason it is a Trend

  • Aircraft manufacturers are constantly looking to develop technologies that will increase efficiency and decrease costs in a highly competitive landscape, and this is driving the trend away from fixed-wing technology toward flexible wing technology. Flexible wing technology has been assisted by developments in advanced materials like graphene and carbon nanotube, which are making wings more flexible and efficient by reducing weight and fuel construction.
  • It is also contributing to the expansion of flight envelopes. Flight envelopes are the capabilities of the design in terms of the load factor, altitude, or airspeed or, in layman's terms, the ability of a single aircraft to be used for multiple purposes, which is beneficial economically.

Machine Learning Skills to Support Trend

  • Machine learning has been used in the development of flexible-wing technology, particularly in the testing and wind tunnel phases, where a range of conditions and configurations are tested. By applying an algorithm to the data collected, AI can predict optimal configurations for different conditions. This is assisting the speed at which the technology is being developed, as well as decreasing the costs associated with development.
  • AI technology also allows for computer modeling that further assists in the development of the technology.

Deficits that Hinder the Trend

  • There is still a considerable amount of experimentation involved in the development of this flexible-wing technology, ranging from the development of materials, through to the configurations of the wing technology. The technology, while seemingly viable, has yet to be fully tested in the real world.
  • Until there is comprehensive testing in the real world, the benefits and advantages the technology offers will be largely hypothetical. It is only when the technology has been proven in the real world, that there is guarantee of the ongoing development and uptake of the technology.

Research Strategy

To determine the current trends in the aerospace industry, we reviewed a range of scholarly research, industry and market publications, and industry experts. By doing this, we were able to identify the two aforementioned trends. We consider something to be a trend if it has generated multiple articles, multiple industry experts are discussing it, and companies within the industry are adopting it.


Part
03
of four
Part
03

Manufacturing - Technology Trends

3D Printing and Nanotechnology are two trends that have the potential to change the manufacturing industry in the same way as the invention of the wheel revolutionized agriculture and industry. The potential these technologies offer is unprecedented. When machine learning and artificial intelligence are applied to the development, the potential grows exponentially. However, like any new technology, they both have deficits associated with them that must be resolved if they are to reach their full potential.

3D Printing

  • 3D printing is an evolving trend that has the potential to change the manufacturing industry. It allows three-dimensional solid objects to be created from a digital file.
  • Objects are created through a process known as layering, in which successive layers of an object are laid down until the object itself is produced.

Reason it is a Trend

  • Recent advances in 3D printing technology, equipment, and materials have seen the costs of adopting this technology become more affordable. As a result the trend has gained traction, as more manufacturers are looking to gain a competitive advantage through the adoption of the technology. With advances in this technology continuing to be made, the trend shows no signs of slowing down in the immediate future.
  • When 3D printing is employed in a manufacturing process, there is scope, not available in a traditional manufacturing line, to change variables easily and with little disruption to the production line. This offers a considerable advantage over traditional manufacturing processes, where changes of this nature can be both time-consuming and costly.
  • 3D printing offers the ability to create complex objects, while using significantly less time and materials compared to traditional manufacturing methods.

Machine Learning Skills to Support Trend

  • Deviations in printing, when printing multiple items that should have the exact same specifications or dimensions, have created issues for manufacturers, due to increased time, costs, and wastage.
  • This has seen significant funding being allocated into research that will alleviate these issues. Researchers are using the information from previous print jobs to train artificial intelligence and create a model that predicts when the distortions are likely to appear and the type of deviation. As more data is fed into the modeling, more precise printing will be possible.

Deficits that Hinder Trend

  • 3D printing allows a range of objects to be produced from a computer generated design. What is not well-known is 3D printing has a high degree of error, with shape distortion being one of the main issues.
  • There are two main reasons for this. Firstly, there are variations between the different printers that are used in the printing process, and secondly, the material that is used to print can be subject to expansion and shrinkage in ways that are not currently able to be predicted by the manufacturer. This means that manufacturers often have to produce several iterations of the product before it is fit for purpose.

Nanotechnology

  • Nanotechnology is still, in many regards, a technology of the future; however, the first generation is already commercially available and is already changing the manufacturing process. This technology allows the manipulation of matter on an atomic, molecular, and supra molecular scale. This enables super-precision manufacturing to a level that has never been seen before.

Reason it is a Trend

  • Modest predictions estimate annual growth of over 30% for nanotechnology in the foreseeable future, due in part to its endless applications. As a result, a considerable investment has been made into the research and development of the technology. This investment has seen the technology become integral across a number of manufacturing industries.
  • It has contributed to safer manufacturing processes and products, improved plant maintenance, created more advanced products, and is becoming an increasingly important factor in the success, performance, and popularity of many manufactured products.
  • To remain competitive manufacturers are increasingly looking to apply nanotechnology to the manufacturing process. The failure to adopt this technology will potentially see those manufacturers lose market share and revenue. This is fueling the trend to incorporate nanotechnology into manufacturing processes.

Machine Learning Skills to Support Trend

  • Nanomatics is a developing area of study. It uses machine learning and data science to explore complex property and structure relationships in nanoscale materials. If nanotechnology is to continue to evolve this research will become increasingly important. The results of this research will inform the development of nanotechnology and its applications across the manufacturing process.
  • Chief Research Scientist, Amanda Barnard of the Commonwealth Scientific and Industrial Research Organization (CSIRO) in Australia explains, "Traditionally structure/property relationships are based on researcher assumption or intuition and then deliberately measured and plotted to confirm a trend. Machine learning needs no input assumption and if a relationship (pattern) exists in the data it will naturally emerge, regardless of whether it was foreseen and targeted in the original series of experiments or simulations."
  • Many of the materials and products that have been developed using nanotechnology require more specific data around structure/property relationships to continue to evolve, meaning machine learning plays a significant role in supporting this trend.

Deficits that Hinder the Trend

  • The potential that nanotechnology offers is unprecedented, but with that potential comes some serious risks, largely because nanotechnology raises issues that are both far-reaching and complex.
  • Many of the risks are associated with the relative infancy of the technology, specifically the ability to measure, predict, and control particles, given the widespread application of technology is currently unpredictable. The long-term implications of nanotechnology are still unclear. There is potential for nanotechnology to cause harm. The wide application of the technology on both a geographical and industry basis amplifies this risk especially as nanotechnology moves from the passive to active stage.
  • Given nanotechnology is increasingly being applied to manufacturing in the food and health industries, this is of fundamental relevance.

Research Strategy

To determine the current trends in the manufacturing industry, we reviewed a range of scholarly research, industry and market publications, and industry experts. By doing this, we were able to identify the two aforementioned trends. We consider something to be a trend if it has generated multiple articles, multiple industry experts are discussing it, and companies within the industry are adopting it.
Part
04
of four
Part
04

Agriculture - Technology Trends

Radio-frequency identification devices (RFID) and blockchain technology are complementary technologies in many respects, with the data that is collected using RFID being recorded on the blockchain. The secure environment provided by blockchain has added to the ability to track products to the farm gate. This has made the supply chain safer and is one of the factors that is driving the current trends to adopt both RFID and blockchain technologies. Other applications include informing indemnity claims made under insurance policies to provide a more accurate assessment of damages and using the data collected to inform "smart farming" technologies.

Radio-Frequency Identification Devices (RFID)

  • RFID identifies an item through radio frequency without any contact. It allows supply chain tracking through the use of an RFID sensor that can be used on both animals and crops. The information collected and stored in the sensor can become a secure and trusted record of the product's history when stored in a blockchain.
  • The devices are being used to manage the supply chain, automate farming processes, manage equipment maintenance, record crop growing conditions, improve efficiency, and manage property access.

Reason it is a Trend

  • The benefits that the technology offers through the collection of growth, weather, and environmental data coupled with the ability to trace products through the supply chain, add value to agricultural produce, which has seen the trend toward the use of RFID increase throughout the agricultural industry.
  • Recent cases of contaminated agricultural products and the need to trace the product to the farm gate have contributed to the growing trend.

Machine Learning Skills to Support Trend

  • Machine learning is increasingly being used throughout the agricultural industry to inform the growing process. The information that is gathered by RFIDs can be input into an AI environment and analyzed to determine optimal planting, growing, and harvesting conditions.
  • As the volume of data available for input increases, the accuracy of the output will improve as the machine learning takes a year-on-year highs, lows, and conditions into account. This makes the combination of RFID and machine learning a powerful tool for the farmer.

Deficits that Hinder the Trend

  • There are still several issues with the technology which may limit its use in practice. There have been ongoing issues regarding the use of the RFID in harsh conditions, especially those with dirt or extreme temperatures, with the units failing to collect the data consistently.
  • The propagation of the crop canopy has resulted in reduced signal strength. This, in combination with a poor range, has resulted in key data not always being recorded. The failure to accurately record the data points has the potential to bias the results when AI is applied. If the trend toward the use of this technology is to continue to rise these limitations must be resolved.

Blockchain

  • Blockchain is a digital register that links its records through cryptography. It provides a secure environment to conduct business without the need for a centralized authority. Records that are created in the blockchain environment are secure and uneditable, due in part to the fact Blockchain was initially created to enable the trade of Bitcoin. Each block is dependent on the previous block in a way that prevents the editing of records.
  • The applications of Blockchain technology have expanded into the agricultural industry, with the benefits of being able to record secure records in a chronological time frame gaining in popularity and applications.

Reason it is a Trend

  • Issues relating to contamination in the supply chain and the inability to quickly trace the source of produce over the last few years, coupled with the increase in the use of RFID, have contributed to the trend toward the adoption of the technology in the agricultural industry. Blockchain technology provides a secure record of the supply chain for products meaning the origin of products can be identified within seconds.
  • Transactions between producers and consumers can be completed on the blockchain without the need for a centralized authority simplifying the process and keeping costs to a minimum.
  • The data recorded on blockchain can be applied to "smart farming" technology that uses AI technology to model and predict various outcomes based on the recorded data. This provides farmers with accurate predictions and modeling that can inform their future decisions.

Machine Learning Skills to Support Trend

  • Insurance in the agriculture industry is fraught with difficulty, especially in relation to losses, where there are multiple assessment methods. There have been ongoing issues around the assessment of damages in indemnity insurance, with a number of productions being unable to be measured. The data that is developed and stored in the blockchain can be used to model various situations through the use of developed AI technology, which will result in a more accurate assessment of damages. Smart insurance contracts that utilize this data are already being used within the industry.
  • The data also has applications in "smart farming," where AI modeling can assist in optimizing farming resources, methods, and resources. By applying RFID data outputs recorded in blockchain to AI technology, output and profitability can be maximized.

Deficits that Hinder the Trend

  • Limitations around the way that blockchain records events limits its applicability. Events are recorded chronological order and time stamped at the time of creation. Once a record is created it is uneditable, which limits its versatility and flexibility.
  • There are issues around the re-creation of recorded data in systems outside of the blockchain environment, so transferring the data to another environment can be problematic. This means that transferring the data recorded by RFID into an environment where AI technology can be applied can be difficult and, in some instances, has limited the use of the collected data.

Research Strategy

To determine the current trends in the agriculture industry, we reviewed a range of scholarly research, industry and market publications, and industry experts. By doing this, we were able to identify the two aforementioned trends. We consider something to be a trend if it has generated multiple articles, multiple industry experts are discussing it, and companies within the industry are adopting it.

Sources
Sources