Our site uses cookies to make it work and to help us give you the best possible user experience. By using our site, you agree to our use of cookies. To find out more about cookies and how you can disable them, please read our cookies statement. 

Cookie Settings

You can manage your cookie settings by turning cookies on and off.

Click on the different cookie  headings to find out more about the types of cookies we use on this site and to change your settings. Please be aware that if you choose to turn off  cookies, certain areas of our site may not work and your browsing experience may be impacted.

For further information on how we use cookies, please see our cookies statement. 

Strictly Necessary Cookies


These cookies are essential for the technical operation of and proper functioning of our site  and enable you to register and login, to easily move around our site, and to access secure areas. Without these cookies our site won't function properly.  

These cookies are required

Performance Cookies

Performance cookies allow us to collect aggregated and anonymous data on how our site is used, such as the number of visitors to our site, how you navigate around and the time spent on our site and also to identify any errors in functionality. These cookies also help us to improve the way our site works by ensuring that you can find what you’re looking for easily, to better understand what you are interested in and to measure the effectiveness of the content of our site. 

Marketing Cookies

These cookies allow us to advertise our products to you and allow us to pass this information on to our trusted third parties so that they can advertise our products to you on our behalf. All information these cookies collect is aggregated and therefore anonymous. No personal information is shared to third parties. Any personal information collected while using our website could be used for direct marketing from Dimension Data only.

Predicting the machine-learning future


From real-time predictions of pro cycling race scenarios, to self-driving cars – here’s what machine learning will offer digital businesses.

We interact with machine-learning systems every day. Without knowing it. From online adverts to the products that Amazon suggests. From the automated sales calls, to the voice recognition systems on our smartphones … these all use machine learning in some way.

We incorporated machine learning and predictive analytics in the pro cycling technology solution we’ve delivered this year. We took historical data, such as past performances of individual riders and previous race outcomes, and combined that with live analysis of race stages. This allowed us to make predictions of race scenarios, such as whether the peloton is likely to catch the breakaway or not.  

This technology is transforming a broad spectrum of industries, even though it may be less obvious than with sporting predictions. In fact, most industries are already benefitting from machine learning. 

Cars, doctors, and schools

One of the most publicly visible implementations of machine learning is the development of self-driving cars. Machine learning is, at its heart, advanced pattern recognition. In the example of self-driving cars, the system is taught to recognise elements of the landscape, such as a stop sign or traffic light, and act appropriately. 

Each interaction with the environment adds more data to the system, allowing it to refine and improve the car’s driving ability. The more information the systems governing self-driving cars have about the way other vehicles, pedestrians, and animals behave, the better the system will become.

Even car manufacturing and servicing will benefit from machine learning in future. In this video, BMW shows how this type of data, gathered from RFID devices, is already helping them keep track of cars throughout their processes. Eventually, machine learning will help to predict and avoid production bottle necks before they occur, improving the services delivered to customers.

Eventually, machine learning may help to predict and avoid car manufacturing problems before they occur

In healthcare, IBM uses the massive computing power of its Watson technology to optimise oncology treatments for cancer patients. The system analyses the available data to find evidence-based treatment options and helps oncologists provide cancer patients with individualised treatment.  

Of course, this underlines the importance of constant and reliable data streams feeding into the system. Watch this video to see how one hospital group, ISPPC in Belgium, uses the Internet of Things to collect medical and other streams of data to create integrated, centralised hospital management. Machine learning will in future add to hospitals’ ability to provide even more intuitive care to patients.

Machine learning may in future help hospital groups such as ISPPC analyse data to provide more intuitive patient care

Machine learning can also enhance the provision of quality education. In the US, 30% of first-year university students don’t return for their second year*. Through analysis of class attendance records, academic performance, and additional demographics, research has shown it’s possible to predict which students are most at risk. Interventions can then be put in place to assist them.

Old school steps up

Some of the greatest potential benefits of machine learning won’t come from Silicon Valley. Rather, it’s coming from the use of the technology by companies that provide services to the industrial, commercial, and agricultural sector, that have access to lots of data which can be leveraged to drive machine learning systems. 

We’ve already seen examples of this with lift manufacturers leveraging sensor data to predict when break-downs are going to occur and schedule preventative maintenance. We’ve also seen farm equipment manufacturers starting to lay claim to the data generated by their equipment, providing them with the data needed to drive sophisticated machine-learning systems. 

For any company looking to get real value out of their investment in machine learning, a few key elements need to be considered: 

  • Firstly, data quality matters. The cleaner the data the better the quality of the intelligence that will be provided. The amount of data matters, as more data will help the system find patterns, but poor-quality data will be counter-productive.
  • Secondly, agility is vital. Machine learning systems don’t exist in a vacuum and, to produce the best results, they need to be constantly tweaked to achieve the desired outcomes.
  • Finally, understand what you want to achieve. Machine learning systems produce their best results when they are set against clear business objectives and a culture of data-driven transformation. Failure to link the investment in machine learning to your organisation’s strategy will result in diminished returns.

*Source: https://arxiv.org/abs/1606.06364

Previous Article: Team Dimension Data: the tech behind the business Next Article: Guest Blog Post - Tech de France

You may be interested in

What running @letourdata taught us in 2018

What running @letourdata taught us in 2018

It takes a team to succeed in the Tour and this applies equally to telling compelling stories from the event

Read blog

Ongoing innovation – how the Tour de France still gets it right after 115 years

Innovation and digital transformation – to stay in business for 115 years, you need more than a strong global brand.

Read blog

Webinar: Meet the tech team bringing the Tour de France solution to life

Join Dimension Data’s Peter Gray and Tim Wade behind the scenes during a live stage of the Tour de France.

Read blog