1: \begin{abstract}

2: 	Multi-output regression problems are commonly encountered in science and engineering. 

3: 	In particular, multi-output Gaussian processes have been emerged as a promising tool for modeling these complex systems since they can exploit the inherent correlations and provide reliable uncertainty estimates. 

4: 	In many applications, however, acquiring the data is expensive and safety concerns might arise (e.g. robotics, engineering). 

5: 	We propose a safe active learning approach for multi-output Gaussian process regression. 

6: 	This approach queries the most informative data or output taking the relatedness between the regressors and safety constraints into account. 

7: 	We prove the effectiveness of our approach by providing theoretical analysis and by demonstrating empirical results on simulated datasets and on a real-world engineering dataset. 

8: 	On all datasets, our approach shows improved convergence compared to its competitors.

9: \end{abstract}

10: