Proposal - sensor_msgs/msg/Contact.msg

[BartlomiejK2]

Description

Simple message for contact sensors, that defines if contact between sensor and some unknown object was detected.

Motivation

I could not find simple message that defines whether there is contact between some defined sensor and other undefined object/objects. Messages like /ros_gz_interfaces/msg/Contact.msg are too complicated and simulator specific for simple contact detection. For contact sensor with force and torque readings geometry_msgs/WrenchStamped.msg would be enough.

Design / Implementation Considerations

# This is a message to hold data from simple contact

std_msgs/Header header      # timestamp is the time the contact was measured
                                        # frame_id is the location of the contact sensor


bool contact                # True if there is contact

Additional Information

I believe that this message could be used for creating future hardware interfaces and broadcasters for ros2_control framework. It also would be helpful for fields where contact sensors are widely used, such as legged robotics.

[Yadunund]

@BartlomiejK2 thanks for the ticket! We've added this topic to the agenda for next week's PMC meeting. You can find the calendar invite for the next meeting here if you would like to join.

[BartlomiejK2]

Hi @Yadunund, thank you for adding this topic to PMC meeting! Has any decision been made regarding this issue?

[ros-discourse]

This issue has been mentioned on Open Robotics Discourse. There might be relevant details there:

https://discourse.ros.org/t/ros-pmc-minutes-for-july-1-2025/45001/1

[mjcarroll]

Hey @BartlomiejK2, we discussed at the PMC meeting this week.

We agreed that a contact interface makes sense for the application areas that you mentioned.

One hesitation is that once messages land in the common_interfaces package, we tend to not iterate on them as much to provide a stable set of interfaces to the downstream consumers. Because of that, it would be helpful to have evidence of this interface being useful "in the wild" before we land it here.

Would you be open any or all of the following?

1. Release this interface in a separate package where people could use it without being part of the common_interfaces
2. Provide some example producers/consumers of the message (hardware drivers, algorithms, etc)
3. Start a conversation on discourse to make sure that other people who would be interested in this interface have a chance to comment?

[ros-discourse]

This issue has been mentioned on Open Robotics Discourse. There might be relevant details there:

https://discourse.ros.org/t/new-contact-message-for-sensor-msgs/45055/1