Managing intermittent RES generation is a major challenge. Connecting households and batteries as VPPs is a valuable building block in that direction. But demand side management is so much more efficient if practicable! We need better market design, smart meters and DSS now.

RT @wolfketter: This is big and can potentially have a large signaling effect! #climatechange - Act now! Judge sides with 16 activists in M….

@KarstenSchroer @INFORMS @TranSciJournal @awscloud @wolfketter @KValogianni @AhadiRamin @markusweinmann @TinglongDai For those who have made it until here: You could be interested in what else is going on at our chair. For other cool research on sustainable smart energy, mobility, machine learning, and human-AI collaboration, check out (19/19).

@KarstenSchroer @INFORMS @TranSciJournal @awscloud Thanks and a big shout out to all the others who contributed to these works @wolfketter @Kvalogianni @AhadiRamin Alok Gupta Micha Kahlen as well as to the other members of the PhD committee @markusweinmann @TinglongDai (18/19).

@KarstenSchroer @INFORMS @TranSciJournal All of us want to thank Karsten for the outstanding work at our chair. @KarstenSchroer, you have been a fantastic colleague, mentor, and friend. Congratulations! You have done great, and we will miss you badly! We wish you all the best for your next steps at @awscloud (17/).

@KarstenSchroer @INFORMS @TranSciJournal The results indicate substantial substitution effects (i.e.: e-scooters reduce bike-sharing). Moreover, thorough analysis of localized treatment effects shows heterogeneity in the effects along different axes like time of day, proximity to the city center and nearby POIs (16/)

@KarstenSchroer @INFORMS @TranSciJournal They analyze the interplay of shared e-scooters and bikes, exploiting large-scale geo-tagged sensor-generated data streams. Causal identification of the effect of e-scooter entry on bike-based micromobility markets is shown with a difference-in-differences approach (15/)

@KarstenSchroer @INFORMS @TranSciJournal 4) Micromobility is a key component of modern urban transport systems. This sector used to be dominated by bicycle sharing schemes. Since 2018, micromobility markets experienced a fundamental disruption following the introduction of e-scooter platforms such as Lime and Bird (14/)

@KarstenSchroer @INFORMS @TranSciJournal In extensive simulation experiments, they show that the method achieves near-optimal EVCH planning results and outperforms alternative candidate solution approaches such as Deep Q-Learning in terms of solution speed and scalability (13/)

@KarstenSchroer @INFORMS @TranSciJournal This approach can greatly improve ex-ante de-risking and decision support in the design phase of service systems such as EVCHs. The method circumvents the need for simplification and problem size reduction, among benefits like more realism, scalability, and flexibility (12/)

@KarstenSchroer @INFORMS @TranSciJournal Circumventing the detail-tractability trade-off of optimization-based methods, they develop a simulated Digital Twin of the EVCH. They then develop an actor-critic RL-framework that interacts with this simulation environment to learn optimal planning configuration policies (11/)

@KarstenSchroer @INFORMS @TranSciJournal EVCH enable daytime charging that takes advantage of high renewable energy production. They are a novel operational system class with cross-system interfaces and a large number of strategic and operational decision variables that result in a highly complex planning challenge (10/

@KarstenSchroer @INFORMS @TranSciJournal 3) EV Charging Hubs (EVCHs) afford high-density EV charging use cases. They are needed as more non-homeowners adopt EVs and as commercial fleets are electrified. Charging opportunities at the workplace, popular destinations such as supermarkets, and fleet depots are needed (9/)

@KarstenSchroer @INFORMS @TranSciJournal Results based on data from the Berlin Car Sharing market show that using the proposed method for repositioning can increase both revenue and profits by up to 3 %pt. when compared to the non-competitor aware model (8/)

@KarstenSchroer @INFORMS @TranSciJournal They determine the value of real-time competitor information in operational decisions by developing an MINLP that returns optimal repositioning decisions considering competitors’ supply; tackling uncertainty by use of an online-learning prediction model (7/).

@KarstenSchroer @INFORMS @TranSciJournal They show how fleet operators can leverage real-time competitor information along with other large-scale urban data sources to make optimal competitor-aware vehicle supply decisions to boost overall market share, utilization and profits (6/).

@KarstenSchroer @INFORMS 2) @TranSciJournal. Shared- on demand mobility is a major trend and balancing vehicle positions is a key topic. Here, competition is often neglected, while there is a finite number of customers. Real-time mobility data is widely available (5/)

@KarstenSchroer @INFORMS Smart sustainable mobility needs careful management and guidance, and they show that IS research can provide this, because of the problem’s complexity and data-driven character. They further develop a research framework comprising seven distinct research opportunities (4/)

@KarstenSchroer @INFORMS Foreseeably, at least a part of the sector will move towards connected autonomous shared electric (CASE) systems. The IS-community is largely silent on this issue, while an uncoordinated CASE mobility rollout could well have detrimental effects on society (3/)

@KarstenSchroer 1) ISR @INFORMS. Modern technologies induce revolutionary transformations of the mobility sector. Yet, most transportation greatly contributes to environmental pollution, while tech-enabled solutions are available or around the corner (2/)