5.5 IBM & SNCB

Through this case study, you'll be able to get a better understanding of how design can be integrated into processes to solve pressing challenges through Big Data, AI and innovation.

Going Beyond Design Thinking

This case study discusses the story of three unique partners who came together to think about how they could create a seamless experience driven by big data, AI, and innovation.

Key takeaways

  • Design Thinking is all about focusing on the user, trying to understand their challenges and identifying the typical user or users.
  • Instead of jumping into solutions prematurely, work with real user stories, not hypothetical users, and start with their needs and how to meet them.
  • AI is something we have turned our attention to significantly in recent years. One of the most common uses of this today is predictive analytics.
  • It is vital to remain open to continuous iterations as you start to learn what is working and where the pain points are.

When Covid-19 hit, IBM and SNCB went into a hyper agile mode or 'accelerator mode', to start working on the real challenges experienced by travellers. They worked through a series of hurdles and quickly concluded that it was important to work on the density of trains and the need to create an interface with passengers that helps them understand how crowded trains were in view of keeping a safe distance and feeling confident to travel again.

Stefan explains that with Covid-19, train capacity went from 100% down to 10% and this required them to start rethinking the way they used to operate. The first 'problems' identified through the design thinking process were "how many people will be on the train?", "Is it going to be full?" and "Will I be able to maintain social distancing whilst on the train?". This was all about enabling customers to make a journey plan safely and confidently. The innovation lab and its team of partners really focused on this as their primary challenge to try and solve.

Focusing on the users, trying to understand their challenges and start by identifying the typical user or users is at the root of the design thinking process. In the case of SNCB, this is something that is core to their innovation principles and so when the crisis hit, they could refer back to personas, adding on top of that so-called key features where you can start to re-draft their user journey from beginning to end. Breaking this journey into stages, we can look into the level of 'achievement' or 'satisfaction' in that journey by asking what is desirable for that customer, what they are looking to achieve and how we can support that.

Once you have a clear idea on that, the innovation process can really get to work. Start by thinking about what data is available and how to use that to support a desirable outcome for the customer.

Speaking with Yves about the role of user experience, he explains that it's a journey for companies to transition from seeing ICT as something that is purely about technology to shifting the mindset towards understanding the importance of user experience and designing technology solutions around it. It is something that IBM has focused on for many years, instead of jumping into solutions prematurely, working with real user stories, not hypothetical users, and starting with their needs and how to answer to that.

Artificial Intelligence is something that we have turned our attention to significantly in recent years. One of the most common uses of this today is predictive analytics. In terms of data, there are two novelties in the case of SNCB. Firstly, the train conductors are part of gathering data and reporting on the situation in real-time, describing the level of crowding they see on a day to day basis. Contextualising where this data is coming from, there is an infinite number of sources which support and enrich the dataset, including weather data through to ticketing data.

There are so many decisions and factors which impact the accuracy and usefulness of the application in this case. Small details, such as the length of the trains or the ability for train attendants to make real-time reports have to be considered when it comes to capturing, analysing and integrating data into a solution to prototype. Yves explains that they decided to launch a 'lighter' version of the main app called 'move safe', in order to prototype it and use observation to see if the accuracy was satisfactory and if the process of gathering, analysing and integrating data was a major priority of the prototyping phase requiring an agile work approach to continuously test, learn and iterate.

In machine learning, we often say 'the more data the better'. Having good knowledge of all the data that can be used is a starting point to building a good AI model that can be used and understanding which data sources can help you to have a good prediction. The approach in this solution is about using a mix of internal data and external data and also ensuring that the data is always available so that the application is stable. The engineering behind the scenes requires a lot of work to bring all the data together, reliably and in a stable environment, which is critical to the success of an AI solution.

Turning to the topic of trust, Ségolène explains that by avoiding private data they are able to avoid additional legal hurdles related to GDPR, which is a further challenge for those working with consumer data to incorporate that into AI models. Considering transparency, it is important to understand how an AI model works, how it takes a decision and ultimately working on a prediction that is understandable by the customer and providing an accurate prediction that can be trusted. A prediction model can never be 100% correct and in the current climate the patterns can make it very difficult to make predictions, as we transition in and out of lockdowns and the situation is evolving all the time.

Operational challenges also come into play here too, for example, the ability of conductors to report all the time whilst also taking care of their core operations. This is why it is vital to remain open to continuous iterations as you start to learn what is working and where the pain points are. Yves points out that even in a single train, there is so much variety because you can have parts of a train which are completely full and other parts which are completely empty and that is where the role of sensors and IoT will ultimately play a key role to synthesise different types of data to produce a more accurate result.

Yves explains that the classic 'big bang' releases just don't work for mobile. There's too much of a void between what is expected by users and what is released by developers. That is not the objective of being agile, this is about reaching a new purpose in cooperating as teams, working towards a minimum viable product (MVP) and then working on iterations from there. This is what happened with Kantify, reaching to a point where randomised data could show interesting results and then setting out further iterations from there. It's about working with deliverables and then correcting them through learnings.

They are working on automated and contactless support in stations using technology but also asking questions related to the business model. At present, train travel is based on season tickets - a subscription-based model - but now the demand is for a different model which offers flexibility, as people mix working from home with working from flexible or satellite offices. This is a huge challenge as they rethink their business model and respond to questions around the so-called 'low-touch economy', the use of virtual assistants, the need for digital ticket sales which incorporate flexibility in a wide range of ways. Beyond these challenges, there are of course many related to the health and hygiene issues which need to be addressed too.

Yves explains that working with innovation and design is certainly a learning curve. Thinking about technology as a process and not a solution requires us to get technology partners involved earlier and there is an increasing trend towards a greater understanding that a mobile website or application is not just a site but its an interface, it is central to the customer experience.When we consider the questions being asked by our sector about whether they should focus on promotion or management, the question is answered by understanding the way in which the world has changed, the way in which technology-driven solutions are central to our real, lived experience. Positioning a brand on image alone means overlooking the role of 'experience' whether that's user experience, visitor experience or the overall lived experience for our customers. Seeing this as a 'nice to have' means overlooking the bigger opportunity to create a real competitive edge by making the customer experience central to the brand experience.

Personalised recommendations or dynamic pricing helps the industry to balance demand throughout the year, which is already helping to balance high-occupancy rates.

Personalisation is also a big opportunity that we have barely touched right now and this will be a really big differentiator for the period ahead.

There are so many exciting opportunities to leverage the potential of technology whether it is Artificial Intelligence, Virtual Reality, Augmented Reality or the huge potential of leveraging the smart use of data. If we consider these technologies in the context of user realities, we can transform the way in which we experience the world around us whilst building incredible experience in spite of challenges.

Going Beyond Design Thinking

This case study discusses the story of three unique partners who came together to think about how they could create a seamless experience driven by big data, AI, and innovation.

Key takeaways

  • Design Thinking is all about focusing on the user, trying to understand their challenges and identifying the typical user or users.
  • Instead of jumping into solutions prematurely, work with real user stories, not hypothetical users, and start with their needs and how to meet them.
  • AI is something we have turned our attention to significantly in recent years. One of the most common uses of this today is predictive analytics.
  • It is vital to remain open to continuous iterations as you start to learn what is working and where the pain points are.

When Covid-19 hit, IBM and SNCB went into a hyper agile mode or 'accelerator mode', to start working on the real challenges experienced by travellers. They worked through a series of hurdles and quickly concluded that it was important to work on the density of trains and the need to create an interface with passengers that helps them understand how crowded trains were in view of keeping a safe distance and feeling confident to travel again.

Stefan explains that with Covid-19, train capacity went from 100% down to 10% and this required them to start rethinking the way they used to operate. The first 'problems' identified through the design thinking process were "how many people will be on the train?", "Is it going to be full?" and "Will I be able to maintain social distancing whilst on the train?". This was all about enabling customers to make a journey plan safely and confidently. The innovation lab and its team of partners really focused on this as their primary challenge to try and solve.

Focusing on the users, trying to understand their challenges and start by identifying the typical user or users is at the root of the design thinking process. In the case of SNCB, this is something that is core to their innovation principles and so when the crisis hit, they could refer back to personas, adding on top of that so-called key features where you can start to re-draft their user journey from beginning to end. Breaking this journey into stages, we can look into the level of 'achievement' or 'satisfaction' in that journey by asking what is desirable for that customer, what they are looking to achieve and how we can support that.

Once you have a clear idea on that, the innovation process can really get to work. Start by thinking about what data is available and how to use that to support a desirable outcome for the customer.

Speaking with Yves about the role of user experience, he explains that it's a journey for companies to transition from seeing ICT as something that is purely about technology to shifting the mindset towards understanding the importance of user experience and designing technology solutions around it. It is something that IBM has focused on for many years, instead of jumping into solutions prematurely, working with real user stories, not hypothetical users, and starting with their needs and how to answer to that.

Artificial Intelligence is something that we have turned our attention to significantly in recent years. One of the most common uses of this today is predictive analytics. In terms of data, there are two novelties in the case of SNCB. Firstly, the train conductors are part of gathering data and reporting on the situation in real-time, describing the level of crowding they see on a day to day basis. Contextualising where this data is coming from, there is an infinite number of sources which support and enrich the dataset, including weather data through to ticketing data.

There are so many decisions and factors which impact the accuracy and usefulness of the application in this case. Small details, such as the length of the trains or the ability for train attendants to make real-time reports have to be considered when it comes to capturing, analysing and integrating data into a solution to prototype. Yves explains that they decided to launch a 'lighter' version of the main app called 'move safe', in order to prototype it and use observation to see if the accuracy was satisfactory and if the process of gathering, analysing and integrating data was a major priority of the prototyping phase requiring an agile work approach to continuously test, learn and iterate.

In machine learning, we often say 'the more data the better'. Having good knowledge of all the data that can be used is a starting point to building a good AI model that can be used and understanding which data sources can help you to have a good prediction. The approach in this solution is about using a mix of internal data and external data and also ensuring that the data is always available so that the application is stable. The engineering behind the scenes requires a lot of work to bring all the data together, reliably and in a stable environment, which is critical to the success of an AI solution.

Turning to the topic of trust, Ségolène explains that by avoiding private data they are able to avoid additional legal hurdles related to GDPR, which is a further challenge for those working with consumer data to incorporate that into AI models. Considering transparency, it is important to understand how an AI model works, how it takes a decision and ultimately working on a prediction that is understandable by the customer and providing an accurate prediction that can be trusted. A prediction model can never be 100% correct and in the current climate the patterns can make it very difficult to make predictions, as we transition in and out of lockdowns and the situation is evolving all the time.

Operational challenges also come into play here too, for example, the ability of conductors to report all the time whilst also taking care of their core operations. This is why it is vital to remain open to continuous iterations as you start to learn what is working and where the pain points are. Yves points out that even in a single train, there is so much variety because you can have parts of a train which are completely full and other parts which are completely empty and that is where the role of sensors and IoT will ultimately play a key role to synthesise different types of data to produce a more accurate result.

Yves explains that the classic 'big bang' releases just don't work for mobile. There's too much of a void between what is expected by users and what is released by developers. That is not the objective of being agile, this is about reaching a new purpose in cooperating as teams, working towards a minimum viable product (MVP) and then working on iterations from there. This is what happened with Kantify, reaching to a point where randomised data could show interesting results and then setting out further iterations from there. It's about working with deliverables and then correcting them through learnings.

They are working on automated and contactless support in stations using technology but also asking questions related to the business model. At present, train travel is based on season tickets - a subscription-based model - but now the demand is for a different model which offers flexibility, as people mix working from home with working from flexible or satellite offices. This is a huge challenge as they rethink their business model and respond to questions around the so-called 'low-touch economy', the use of virtual assistants, the need for digital ticket sales which incorporate flexibility in a wide range of ways. Beyond these challenges, there are of course many related to the health and hygiene issues which need to be addressed too.

Yves explains that working with innovation and design is certainly a learning curve. Thinking about technology as a process and not a solution requires us to get technology partners involved earlier and there is an increasing trend towards a greater understanding that a mobile website or application is not just a site but its an interface, it is central to the customer experience.When we consider the questions being asked by our sector about whether they should focus on promotion or management, the question is answered by understanding the way in which the world has changed, the way in which technology-driven solutions are central to our real, lived experience. Positioning a brand on image alone means overlooking the role of 'experience' whether that's user experience, visitor experience or the overall lived experience for our customers. Seeing this as a 'nice to have' means overlooking the bigger opportunity to create a real competitive edge by making the customer experience central to the brand experience.

Personalised recommendations or dynamic pricing helps the industry to balance demand throughout the year, which is already helping to balance high-occupancy rates.

Personalisation is also a big opportunity that we have barely touched right now and this will be a really big differentiator for the period ahead.

There are so many exciting opportunities to leverage the potential of technology whether it is Artificial Intelligence, Virtual Reality, Augmented Reality or the huge potential of leveraging the smart use of data. If we consider these technologies in the context of user realities, we can transform the way in which we experience the world around us whilst building incredible experience in spite of challenges.