© 2022, The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature.Paralysis caused by physical trauma is a common disease today, with approximately 30% of paralysis caused by this trauma. The disease in question both physically restricts mobility and brings along psychological problems. Especially in advanced ages, paralysis becomes much more difficult and requires serious care since it causes many effects in elderly people’s daily routines, required specific healthcare services, costs…etc. Therefore, in this study, it was aimed to design an Internet of Health Things (IoHT)-based unmanned robot vehicle for paralyzed patients, by using deep learning for control of the system and creating a connected healthcare synergy. In this context, the electronic part of the robot was designed first and then a deep learning-based model was created. Before creating the model, the data obtained from the sensors were adequately pre-processed, and the related problem formation was shaped for a Deep Learning process. In order to compare the Deep Learning model with other machine learning techniques from the state of the art, different models based on Random Forest (RF), Decision Trees (DT), Naive Bayes (NB), Support Vector Machines (SVM), k-Nearest Neighbors (k-NN), Logistic Regression (LR) and Artificial Neural Network (ANN) were generated using the same data. The performances of these models were compared and the obtained results were analyzed. According to the experimental results, the Deep Learning model has the highest performance with 99.5% success rate and 0.5% loss rate. This work presents some relevant results in which using deep learning techniques point excellent performances in the area of IoHT applications. A second contribution of this paper is related to the evaluation of the communication flow of the systems, to analyze if it was good enough to cover a smart environment. The findings in this manner were positive in terms of technical analysis and the patient feedback taken from a total of 10 paralyzed patients.