Sorting objects or products manually in industries using human labour consumes time and energy, especially when it involves large volume of products. Human sorting operations could be replaced with robotic implementation, which would save time and effort while producing better results. Robotic arm is a type of programmable mechanical arm, almost similar to the human arm which is programmed to faithfully carry out repetitive actions with a high degree of accuracy. However, creating an artificial sense of touch that is comparable to human's has been challenging. One of the challenges is the detection of objects with different colours. In this work, a light intensity to frequency converter method is used to detect colour. The robotic arm is controlled by a DC servo motors microcontroller-based system. An IoT platform is being used to monitor the detection process, especially the counting for each sorted colour block. Furthermore, the study also performed a comparison of the robotic arm indoor and outdoor efficiency. The chosen TCS3200 sensor operates by distinguishing different generated frequencies for different light intensities. It could distinguish the colours at 95% in a bright condition and 91% in a dimmer condition, based on object detection at 5 cm away from the sensor.