North East Innovation Meetup
Welcome - Colin Eberhardt, CTO - Scott Logic
Colin welcomed those there to the North East Innovation Meetup, Scott Logic was founded twenty years ago to help various organisations and tackle complex software. Scott Logic have well over one hundred employees is the business and are proud to be in the North East.
Digital Transformation: Empowering the Masses - David Forrest, Manager - AMIEO Digital Manufacturing at Nissan
Nissan make vehicles here in the North East including the Nissan Juke and the electric car the Nissan Leaf and have recently produced the 500,000th car there. Nissan employee over 6,000 people with 35,000 people supported with two production lines and will have an all electric version of the Nissan Qashqai. The production involves just in time delivery for parts and the systems to manage this are quite significant, they have a new battery plant which is one of their suppliers.
David's team has a mission and strategy to drive digital transformation across AMIEO to improve cost, quality and delivery and the first strategy is to be data driven with a single source of truth so they can apply visualisation and machine learning an AI on these data sets to help be digitally enabled and people centric with citizen development and digital literacy is very important from all levels in the business and elevate them with the tools they have and the things that they can do.
They have four teams working together including digital manufacturing data, digital technology, digital quality and supply chain transformation including how to ensure machines they buy will work first time with use of AR and VR. They will also prevent and predict issues by applying machine learning and supply chain transformation with a hands-off supply chain and visibility of supply chain including where is everything in terms of that supply chain.
This is innovation for them in their sector but may not seem like it to others, it is easier in a greenfield business but they definitely have innovation for them. They want to have right first time and safety with digitally enabled manufacturing they also do real-time quality reporting with dashboard so are looking at ways to skill up with XR and they do as much as they can to reduce energy and use machine learning to predict energy usage to purchase electricity at a cheaper price and get rid of silos that exist in their business.
They have installed a data flow platform for a single source of truth that can be exploited with an ETL platform using open-source to connect to any data source in the factory from a PLC or Robot or SQL database along with an Excel file on someone's computer and they push data into two places with a cold-data store with Snowflake in AWS and have a hot data on-prem database using SQL.
Is there an internal drive for this or did they see others doing this? They realised they were inefficient with their data so worked together with Renault to improve efficiency in their data and looked at architecture so it can be replicated at other plants around the world.
The problem and opportunity is the tip off the iceberg with their industrialised data but there is non-industrial data including 10,000 Access databases along with paper and manual processes so need to get everyone involved and get their data connected to their platform from an Access database and do it properly. They are launching citizen development who is anyone in the company who can get training and access to data along with support from their team and enabled a toolset including PowerApps, Power Automate, Power BU and tableau plus have training and skills both in person and digital with Udemy and LinkedIn Learning and had leadership buy in and make sure that data was secure and goals were clear and can keep iterate and improving what you do with that data along with supporting people and establishing robust governance and support.
Self-service data analytics is a key one to allow people to access data they are curating through the platform where they connect PLCs, IoT sensors and push this into a cold and hot data store and have various platforms including AI and Machine Learning and are constantly growing their tools with data sources from systems or sensors using LoRaWAN to ensure data is high quality and trustworthy. They are improving everyone's digital literacy with open day events where they invite members of staff to get engaged, send newsletters and have digital champions across the business and have people. They also do regular webinars and have hundreds of people dialling into these to see what they are up to and have curated a lot of information into their SharePoint site and communications is one of the most important parts along with holding regular workshops to upskill a lot of people in digital tools.
Made North East is a regional and national centre of expertise focusing on enhancing skills and innovation in manufacturing excellence, electrification and digitalisation which Nissan is doing but is open to any manufacturing company to train in any automation and electrification and looking at what they can offer at a digitalisation point of view from kids to experts at the top of a company at different levels including apprentices who will be sent there. This will help companies understand their digitalisation strategy including how to get access to funding, skills and training but want the centre to show best practice and share knowledge along with a fully accessible sandpit environment to try out things and tools. They are also considering advanced logistics including investment in North East as an AI growth zone, there is other technologies they can put in there with the centre to demonstrate this stuff and is relatable and people can appreciate machinery which is similar to theirs.
Questions
Creating a data lake and single source of truth is something Scott Logic do, the first thing is hoovering up the data, but problem can be identifying the single source of truth and is not clear what should be this? They have people who have an Access database which is a copy of data so they try to go to the source of the data so have connectors that talk to PLCs and have a data engineering team to look at the root of the data.
How did they get onto this? They had to persuade their bosses and gave grown the team from one to over twenty people and have proven it is worthwhile but the biggest challenge is change management, they can create the best tool but getting people to use it is the biggest problem but can be communication, education and buy in from leaders and have done a workshop with managers but there are some people you can't change the minds of, 20% will embrace it, 20% will dismiss it so the focus is on the middle.
What has it actually given to the organisation? They have to justify the teams existence and budget, they track every use case and it has to have a benefit and is then financed and they have procedures to track this, the benefits need to be cost reduction, cost avoidance including quality and none value added time reduction to make people's jobs more efficient but finance doesn't recognise it as have take time from their task to focus on something else is not possible to quantify.
Creating a culture for AI innovation - Molly Pace, Business Development Manager - Scott Logic
Molly talked about AI innovation from a people focused point of view as people are Scott Logic's best asset. Don't forget the human in human-in-the-loop, AI is automating tasks not people it is replacing and automating tasks and this means the most successful applications of AI still have the element of human-in-the-loop. We are forgetting the human part, we talk about the tech and outcomes and how successful the outcomes are so we are not assessing the impact on humans. We are now collaborating with AI and it is crucial to assess the impact and the culture of people who are working with AI. They are focusing on the neuroscience and how AI is impacting our reward system and motivation to do certain tasks and what this means for AI strategies for companies.
Motivation, there is intrinsic and extrinsic motivation. Intrinsic motivation is something you want to do and have a curiosity to do so and extrinsic is where there is an external reward or factor you want to avoid. With running people may be intrinsically motivated to run or may be extrinsically motivated to run in a race in Newcastle that has a map of Sunderland on with the Great North Run. There will be tasks that you do because you enjoy them but there are task they do to avoid your boss coming over and being angry at you. You can think of things that you are intrinsically motivated to do and things that are extrinsically motivated to do but are more likely to do and progress with the former.
AI's influence on intrinsic motivation is that people found that while AI helps people complete tasks it also decreased long-term motivation and psychological wellbeing as weren't getting the same sense of reward out of it and in turn is was reducing their wellbeing. With developers there is a concern that it will take away the best part of their jobs and that AI will get all the good bits and take creative parts and developers will be left doing things like code reviews because they have to do anyway. There has been research in game development where people were asked to use AI in the creative parts of their role but this reduced the sense of wellbeing and reward and it created long-term problems later on. To support AI innovation we need strategies that take into account the impact our AI strategies will have on them.
Designing an AI strategy that supports intrinsic motivation is to use AI for tasks that don't take away ware people aren't intrinsically motivated to do, we shouldn't set a blanket policy without explaining why it is helpful - why we are doing something and what impact it is going to have. Protect the task and skillsets that people attribute sense of self to such as asking a writer to use AI to help write may not be helpful and think beyond short-term gains.
They have worked from an idea to a production ready chatbot which was a culture of experimentation enabled innovation and empowered autonomy for motivation so had a purposeful application that addresses real internal frustrations which helped customer service agents resolve tickets faster by retrieving and summarising relevant information from multiple sources and they found that feedback from users that over 80% of the chatbot answers were useful in closing tickets and helped improve productivity to solve a pain point they already had and addresses a real frustration with something they weren't intrinsically motivated to do.
They also did some work with which did it in a nice way which was a legal tech company which came off the back of a hackathon to develop a generative AI powered microservice that extracts key legal events from complex documents to help legal professionals work faster without compromising accuracy to create a timeline of where documents were relevant to a legal case which was being given to summer interns so thought why not get generative AI to do this instead so they can do more learning when they are there. Their culture created that space for hackathons and to experiment to help and not replace and they did a lot of user testing throughout to get trust and validation to use the system. They had a startup mentality that allowed them to innovate and give space to fail and test things out to create something really helpful in the end.
Some things you can start, stop and continue to do in terms of AI strategy. Start to empower experimentations to try things out, support autonomy and purpose and create collaborative environments for different perspectives on things. Stop mandating adoption without context, don't undermine expert roles like telling writers to use code for all their writing so don't undermine expertise and make things that help them and the other thing is stop ignoring the cultural shift, there is nothing physical that is changing with paper to computers it is more obvious so with this we are not recognising the shift that is changing as nothing physically is changing. Continue to celebrate learning not just success along with continue to listen and implement feedback.
Questions
AI is now contributing and reducing ability for people to critically think, they were getting Copilot to write Git commits and thought about wanting to do this themselves? Some things that are a little bit challenging is the sense of reward afterwards and making things easier is losing that reward.
Which point do you move beyond motivation, you can make everyone more efficient so then how can you productise things or make money from these, how can you move beyond productivity gains? There is a lot of product things out there, the productivity gains are easy to see. This can feel like a first phase but get into that phase where people can feel threatened? You have to probably think about what it is coming for some things at some point and AI is mostly used for productivity gains and certain sectors will see things differently and may be more of a pivot to see where are focusing rather than a complete change like with paper to computers, people would have thought the same thing that they were going to be replaced by a computer.
Think about the impact on commercial or financial perspective, can look at things that people are doing but people can become cynical about things that are being pushed there way and may reject things as there isn't the credibility with what is coming your way as you are being bombarded with things.
Technology & Innovation: North East of England - Iain Robinson, Global IT & Service Director - Crane NXT
Iain talked a little about their company which has been in the North East for a while and want to tell people about the exciting things they are involved with. Iain has worked with IT for their whole career. Crane empowers brands and governments to make confident decisions in the moments that matter including payments and innovations with vending machines and also have a very big currency business unit that actually makes the US dollar and Iain came from De la Rue which used to make currency in the North East and also used to make passports and this moved abroad. Crane is in the same business and spun off a whole business for authentication which was carved out and acquired from De la Rue and have been exposed to new technology and cool manufacturing stuff.
They are proud of their North East roots and are based on Gateshead quayside and the Gateshead team includes global IT, service management, software development, project management and physical development along with R&D. They have a Washington team which includes material science, manufacturing design, factory operations and customer care. They authenticate products including product licencing or if a brand being sold somewhere they haven't authorised and when have a brand you can track it through its lifecycle and have 360 protection for those brands. Their biggest market is in the middle east, they didn't previously pay taxes on things but now they pay taxes so need to be able to track products and be traceable along with identity verification.
Physical products and identity including holograms and micro-optics for brand protection which can include Newcastle United tops which are very tough to replicate to ensure something is an authentic product but isn't track and traceable. Software development is the data-driven brand protection where they do leverage AI and big data analytics on where their products are being sold and real-time insights along with identification for governments. 70% of Mastercard / Visa holograms are made, developed and produced in Washington in Tyne and Wear with over 2.28 billion made every year as card fraud costs $30 billion annually but cards are becoming a more legacy product with payment systems such as Apple Pay as it is a business which is liable to forgery. The hologram is often the only brand mark on a credit card and if cards don't have holograms, then any counterfeiting won't investigated by the US authorities.
Overt brand protection can include trim, laminate, labels, films and foils which can be incorporated into packaging and hangtags like for NFL merchandise along with data protection pages for Australian passports being done in the UK. They work on many overt security products and form factors and have features such as fresnels, contrast switch, pure image with animated colour, moiré effect to help people look at something and know it is legitimate.
They have a solutions business to create physical markers for products along with a digital Id to track a product throughout the lifecycle including registration and validation, physical marks and more such as tobacco manufacturer who would order stamps to be applied and then could send messages to their system where border guards or end user could scan a code with their phone to see if a product matches or is legitimate and how old the product is and they supply this to many businesses mostly in the Middle East. Their IT teams look after these, and the software developers create the systems for an end-to-end system for government revenue. Middle East also wanted every single stamp to be unique from each other with additive finger printing to make every item unique and have only customer in the world who does this and even with high value printing presses would be impossible to fabricate these which also include a unique digital ID which is uploaded and validated to their cloud system which allows it to be cross referenced to verify authenticity using AI and machine learning to harden experience and therefore becomes more robust over time. Machine learning is more about environment of where the label is scanned, on any device and to try and make this successful you need to use learning to make it easier.
What's coming - innovation for the future with inclusivity in the North East, it is essentially a brave-new digital world with layers of authentication with more layers being added and moving into the quantum age and need a strategy that puts authentication into hands of consumers where they are being pushed things rather than pull. Create non-additive authentication where scan fibres on clothing or other unique secure signatures. Digital product passports about the whole cycle of a product including the resources it took and it being sold and recycled and used again so their labels need to be part of this. It is all about data with deep market intelligence along with sustainable scalable path to solving global counterfeiting.
Questions
It is about proving physical things are authentic so what about digital products which are much easier to fake such as videos of world leaders? They are involved in conversations about this but they are physical security printers but there is a market for physical digital link with a digital anchor to marry physical product to digital data to track and trace it to see where brands are bring sold and used.
Are you in talks for using where any product is in terms of recalling as if know where something is? This isn't something they do but they do enable companies to do things with their solutions to enable these scenarios and is part of their value proposition for deep value intelligence and have all this data and this data is valuable so how do you extract that value and this is something brands and customers pay for.