This paper proposes a new system to translate an Arabic Sign Language (ArSL). The system consists of two sub-systems: the first, Speech to ArSL translation Subsystem. This sub-system is mainly based on the speech recognition engine. The second is ArSL to speech translation subsystem to translate the images of signs into speech, mainly based on GTTS library. This system will be easily used by both groups (the hearing disabled and normal persons) who want to learn ArSL and will support more communication between them. Among the advantages of this system is that it is expandable by adding new signs to the database and can be accessed online at: https://sr.gravita-demo.com/. The proposed system was evaluated using several methods. The findings show that the system can translate ArSL with a recognition rate of 99%. An experimental approach was used to evaluate the effectiveness of the proposed system. Results show that the proposed system was effective and the ability of Arab deaf students to learn improved significantly. In addition, the system's performance evaluation questionnaire revealed that system users preferred this approach to ArSL learning and teaching and acquired new concepts, which predicts a promising future for this system in academic environments.