Accurate in-situ stress characterization is essential for predicting subsurface responses to interventions such as underground construction and fluid injection or extraction. At depths of 2–5 km, the stress field is heterogeneous and influenced by complex geological features, making reliable stress measurement both technically challenging and operationally critical. Borehole stability is another key concern, as deep boreholes are prone to stress-induced deformations such as breakouts that can damage equipment, impede drilling, and even lead to failure. This talk will present an improved technique to estimate the full stress tensor from data recorded with the SIMFIP probe. By combining three-dimensional displacement measurements during fluid injection with pressure transient analysis, the method enables stress tensor estimation from a single test. I will present a complete stress profile obtained along a sub-horizontal borehole at the Bedretto Underground Laboratory (BedrettoLab) in Switzerland, revealing significant stress heterogeneity and rotations around an intersected fault zone. I will also present the development and field deployment of a novel borehole probe to study thermal breakout initiation. The probe can heat a packed-off borehole section while measuring borehole wall displacement. After extensive calibration in the laboratory, the probe was deployed in the BedrettoLab, and three in-situ heating tests were successfully conducted up to 140 °C. These experiments will advance understanding of thermo-hydro-mechanical coupling at borehole walls to study breakout initiation and borehole stability. Finally, I will briefly introduce the first CO2 injection pilot (CiTru) planned in Switzerland, which is evaluating the feasibility of injecting CO2 into Muschelkalk limestones at depths over 1000 m.