DIGIT’s Intelligent Automation Virtual Summit brought together a range of speakers to discuss issues in artificial intelligence (AI) and automation.
Now in its second year, the conference looked at how the Internet of Things (IoT) will transform factories and cities, different uses for AI and the ethical challenges arising from increased automation.
In the opening session of the day, delegates heard from Nvidia Technology Evangelist Sarah Mannion, Intelligent Growth Solutions CEO David Farquhar and Larissa Suzuki, Associate Professor at UCL.
Mannion detailed how Nvidia’s technology is finding its way into smart manufacturing facilities. It is AI that bridges the gap between automated processes and smart processes, where the machines gather data from their surroundings, use it to make decisions and, crucially, communicate with other machines.
For example, Nvidia has been working with online grocery shop Ocado. When picking produce, Ocado’s machines need to understand the difference between soft and damageable products like fruit and hard and heavy products when filling a pallet. It is AI that helps them understand that.
IGS Chief Exec David Farquhar talked about his company’s work using AI in vertical farming. The technology can tailor the micro-climate of individual trays across an entire stack to ensure greater yields and ensure reduced resource demand. AI is needed to process the vast amount of data involved in simulating the weather to create a micro-climate – the wind, the sun and the rain.
Crucially, vertical farming means that farmers are no longer at the mercy of the weather. With the UK facing a major wheat shortage and the US in a similar situation with wild blueberries, this could prove vital for food sustainability.
With farming as the world’s most polluting industry, finding newer, more efficient ways to work is vital to protecting the planet, Farquhar told delegates.
- How Smart Cities Are Fighting the Covid-19 Pandemic
- Smarter Grid Solutions R&D Projects Surpass £20m Milestone
Continuing on the theme of sustainability, Suzuki gave delegates a glimpse into the future of smart cities.
Increasingly, Suzuki asserted, urban areas will present “immense” problems for people and the environment. Each year, urban environments are responsible for $270bn in energy waste and $173bn in food waste, while one-quarter of cities globally face issues with water scarcity and congestion.
One of the most concerning aspects of modern city life, Suzuki explained, is the extent of pollution. More than 90% of the world’s population is exposed to concerning levels of pollution, and with 2.5 billion people expected to move to cities in the coming decades, new technologies could help cities work more efficiently and improve sustainability.
With millions of sensors, information across a city can be gathered and utilised to produce valuable insights and ensure resources are used to maximise positive impact for citizens.
All three speakers in the first session discussed the increasingly crucial role of edge computing in smart environments. Instead of processing data at a centralised facility, data is processed close to the user. This is important in time-sensitive processes, such as predictive maintenance and self-driving vehicles, where a round trip to a data centre can add precious milliseconds to the work.
Session two followed with presentations from David McEnaney, VP at Blackrock’s Process Automation Centre of Excellence, Nokia’s CTO of Global Enterprise UK&I Steve Evans and Chief Scientific Advisor & Co-Founder at Space Intelligence, Ed Mitchard.
McEnaney explored robotic process automation and how it can be used to simplify processes or reduce costs. Similarly, he underlined the crucial role of ‘citizen engineers’, where programming tools are designed specifically to be used by non-programmers.
This, McEnaney said, helps ensure that different perspectives go into creating programmes and provides workers with valuable new skill sets.
Evans, meanwhile, talked about Industry 4.0 and the potential it holds. With numerous physical industries lagging behind the ICT sector in terms of investment, Industry 4.0 can provide them with means to increase productivity and efficiency. He estimated that IoT technologies could add between $3 trillion and $11 trillion of economic value globally by 2025.
He alluded to the advent of 5G technology and the use of radar-based technology to promote data flows to get intelligence from intelligent machines in real-time. Evans also noted the importance of 5G in mission-critical processes, such as preventing accidents in the event of a failure, as well as using 5G to remove the need for wiring up machinery, making factory spaces more adaptable.
- Scottish 5G Groups Join Forces to Accelerate Roll-Out
- Anti-5G Group Crowdfunds More than £100,000 to Halt UK Roll-Out
The final presentation of session 2 was Ed Mitchard from Space Intelligence. His company is using AI to analyse satellite data to fight climate change. With thousands of satellites creating hundreds of terabytes of free data, processing and utilising that data is an intensive business.
By using AI, different kinds of information, such as optical, synthetic aperture radar, gravity and lidar can be combined to produce more accurate and more informative images.
With a well-written algorithm, AI can then be used to pick out important data from the images. Space Intelligence uses that information to track forests, peatland and other geographical features in the fight against climate change.
Ethics and AI
The final session of the day saw ICO’s Principal Technology Adviser Andrew Paterson, Director of Global eHealth at the University of Edinburgh and Chair of the National Expert Group on Digital Ethics Claudia Pagliari and Ethical Intelligence CEO Olivia Gambelin discussing the socioeconomic challenges of automation and how they can be met.
Paterson discussed the legal implications of automation and cybersecurity. He noted that, while automation is a powerful tool, it cannot do everything – there were always be a need for human involvement.
Automation is vital to scaling networks – manual measures may work for small-scale networks but are unfeasible for managing massive networks since human beings are simply not efficient enough.
However, GDPR article 22 protects humans from being subject to purely automated decisions – we have the right to be handled by humans and that demands human intervention and transparency.
- Report | Employee Mistakes Cause Almost Half of Cybersecurity Issues
- IoT Programme Secures Four SME’s for Cybersecurity Innovation
Claudia Pagliari from the University of Edinburgh looked at the importance of integrity in cybersecurity systems – the need for them to be secure, reliable and properly managed.
However, she noted, integrity is bigger than that – it includes the ethics of data protection. Personal data cannot be treated in a secretive manner and the algorithms that cover them must be open to examination. She touched on the recent grading fiascos that have hit both Scottish and English schools, as children find their grades and futures determined by opaque and confusing algorithms.
Finally, Gambelin looked at the ‘triple bottom line’ of the ethical usage of AI; people + planet + profit. For an AI to be ethical, all three elements need to be present.
When one of the steps is missing, the whole system suffers, and if one is more important than the other, the whole system can fail.