Innovation in the Age of AIChallenges

Legal Tech

How might we support business professionals checking during contract review, if all required documents exist and all relevant topics are addressed?

Contracts are an important legal cornerstone of the global economy. Especially in high volume transactions contracts can get very extensive and complicated, and reviewing these contracts is time consuming, burdensome and prone to error. In most cases, before a contract can be signed it is necessary to review if all required documents exist and all relevant topics (legal provisions) are addressed. Today, this is mostly done manually. Rule-based and other traditional technical approaches have proven to be impracticable. Is it possible to develop an assistant based on deep learning methods to help speed up and facilitate this task?

 

Automation

How might we improve conditions for factory workers in production and increase quality through automation?

Employees in manufacturing companies still experience a strenuous workday. Whether it's lifting heavy loads or monotonously inspecting a specific workpiece, fewer and fewer people want to do these tasks. Instead, automation is moving onto the scene: machines and robots, often guided by industrial image processing, take over these tasks. Humans take care of monitoring, optimizing and maintaining production and software. What can this process look like to enable a better work situation while maintaining high quality results? What possibilities does AI offer in automation, especially in machine vision, to master this future task? And how can this process also be successfully designed for assembly line and production workers?

Dr. Christoph Garbe is CEO of HD Vision Systems as well as a physicist with a doctorate and post-doctoral degree specializing in image processing. With his start-up, he wants to help manufacturing companies make their production fit for the future using 3D light field technology and AI. 

Healthcare

How might we empower patients to work together with healthcare providers and pharmaceutical companies to improve individual treatment options and outcomes?

A lot of medical data is generated throughout a cancer treatment and collected and documented by various tools and systems. This data harbors an immense potential to learn about optimal therapy sequences and real life factors influencing - meaning harming or benefiting - successful treatment. Is it possible to develop an AI-backed health data network for optimal treatment decision making?

 

(Institut für Bildungswissenschaft, Universität Heidelberg & Pädagogische Hochschule FHNW)

Education

How can we make students' learning visible to themselves and thus support self-regulated learning processes as well as successful graduation?

New at university, the first exams approach – can I manage that? Just before graduation: the last exams, will everything go well? Learning is challenging. Especially at the beginning of your studies, when everything is new, and at the end, when many exams are coming up. This is where good planning and organization, choosing appropriate strategies, and perseverance are important. It is also helpful to keep an eye on your learning progress and adjust your approach if necessary. These skills are referred to as “self-regulated learning (SRL).” Studies show that SRL is an important key to successful study. For many students, however, SRL is a challenge.

This is where our challenge comes in: We want to support students' SRL by making their learning processes visible to them and provide a basis for managing their learning during exam periods. 

Silke Hertel is a professor at the Institute of Educational Science at Heidelberg University. Her research focuses on SRL of children, adolescents and students. Yves Karlen is professor for educational-psychological teaching and learning research at the University of Teacher Education FHNW. His research interests lie in the promotion and teaching of SRL in schools and universities. Both are very interested in innovative methods for diagnosing learning processes.