Malicious object detection in 3D printer device management
Systems and methods for 3D printer management can allow or reject printing of an object based on a model that is trained with machine learning. In one example, the model classifies the object according to object type. The object type can be compared against a list to determine whether to block the object from printing. The object can also be compared to previously printed objects. The object can be blocked from printing if a combination of the object and a previously printed object amounts to a threshold amount of a blacklisted item. The lists can be specific to users, such as an organizational group to which the user belongs. A print server can apply the model prior to forwarding the object to a 3D printer for printing. Both the models and the lists can evolve based on machine learning, such as based on which print decisions receive override from administrators.
1. A method for 3D printing management, comprising:
receiving, at a print server, a request to print a first object at a 3D printer, the request being associated with a first user profile and including an object data file of the first object, the first user profile being associated with a group identifier;
applying a first model to the object data file, wherein the first model outputs a first classification of the first object based on shape information of the first object;
predicting, based on the first classification, that the first object is a first component of a blacklisted item;
applying a second model to the first object using a data table with a print history of user profiles associated with the group identifier, wherein the second model outputs a second object that is a second component of the blacklisted item, the second object having been previously printed using a user profile associated with the group identifier other than the first user profile;
determining that the second object is a whitelisted item;
based on the determination that the second object is a whitelisted item, reducing a percentage that the second object counts toward the threshold amount of the blacklisted item by a predetermined amount;
determining whether a combination of the first object and the second object represents more than a threshold amount of the blacklisted item;
denying the request to print the first object if the combination of the first object and the second object represents more than a threshold amount of the blacklisted item; and
allowing the request to print the first object if the combination of the first object and the second object does not represent more than a threshold amount of the blacklisted item.
2. The method of claim 1 , wherein applying the second model to the first object includes using a data table with a print history of a second user profile, wherein the second user profile is linked to the first user profile based on shared communications between the first user profile and the second user profile.
3. The method of claim 1 , wherein:
applying the first model includes applying a tag to the first object based on a determined classification, the tag corresponding to the blacklisted item, and
applying the second model includes determining that the second object also includes the tag and determining an amount of the blacklisted item that the combination of the first object and the second object represents.
4. The method of claim 1 , wherein a print history of the first user profile is stored in the data table, wherein the print history includes a history of objects printed using the first user profile.
5. The method of claim 4 , further comprising comparing the data table to a blockchain record of user print histories to verify that no records have been removed from the table.
6. The method of claim 1 , further comprising sending an alert to an administrator that identifies the combination of objects used to deny the request.
7. The method of claim 1 , wherein:
the first model and the second model are trained using machine learning; and
the method further comprises:
retraining at least the first model or the second model based on a determination whether a combination of the first object and the second object represents more than a threshold amount of the blacklisted item.
8. A non-transitory, computer-readable medium containing instructions that, when executed by a hardware-based processor, performs stages for 3D printing management, the stages comprising:
receiving, at a print server, a request to print a first object at a 3D printer, the request being associated with a first user profile and including an object data file of the first object, the first user profile being associated with a group identifier;
applying a first model to the object data file, wherein the first model outputs a first classification of the first object based on shape information of the first object;
predicting, based on the first classification, that the first object is a first component of a blacklisted item;
applying a second model to the first object using a data table with a print history of user profiles associated with the group identifier, wherein the second model outputs a second object that is a second component of the blacklisted item, the second object having been previously printed using a user profile associated with the group identifier other than the first user profile;
determining that the second object is a whitelisted item;
based on the determination that the second object is a whitelisted item, reducing a percentage that the second object counts toward the threshold amount of the blacklisted item by a predetermined amount;
determining whether a combination of the first object and the second object represents more than a threshold amount of the blacklisted item;
denying the request to print the first object if the combination of the first object and the second object represents more than a threshold amount of the blacklisted item; and
allowing the request to print the first object if the combination of the first object and the second object does not represent more than a threshold amount of the blacklisted item.
9. The non-transitory, computer-readable medium of claim 8 , wherein applying the second model to the first object includes using a data table with a print history of a second profile, the second profile being linked to the first user profile based on shared communications between the first user profile and the second user profile.
10. The non-transitory, computer-readable medium of claim 8 , wherein:
applying the first model includes applying a tag to the first object based on a determined classification, the tag corresponding to the blacklisted item, and
applying the second model includes determining that the second object also includes the tag and determining an amount of the blacklisted item that the combination of the first object and the second object represents.
11. The non-transitory, computer-readable medium of claim 8 , wherein the print history of the first user profile is stored in data table, wherein the print history includes a history of objects printed using the first user profile.
12. The non-transitory, computer-readable medium of claim 11 , the stages further comprising comparing the data table to a blockchain record of user print histories to verify that no records have been removed from the table.
13. The non-transitory, computer-readable medium of claim 8 , the stages further comprising sending an alert to an administrator that identifies the combination of objects used to deny the request.
14. The non-transitory, computer-readable medium of claim 8 , wherein:
the first model and the second model are trained using machine learning; and
the stages further comprise:
retraining at least the first model or the second model based on a determination whether a combination of the first object and the second object represents more than a threshold amount of the blacklisted item.
15. A system for 3D printing management, comprising:
a memory storage including a non-transitory, computer-readable medium comprising instructions; and
a computing device including a hardware-based processor that executes the instructions to carry out stages comprising:
receiving, at a print, a request to print a first object at a 3D printer, the request being associated with a first user profile and including an object data file of the first object, the first user profile being associated with a group identifier;
applying a first model to the object data file, wherein the first model outputs a first classification of the first object based on shape information of the first object;
predicting, based on the first classification, that the first object is a first component of a blacklisted item;
applying a second model to the first object using a data table with a print history of user profiles associated with the group identifier, wherein the second model outputs a second object that is a second component of the blacklisted item, the second object having been previously printed using a user profile associated with the group identifier other than the first user profile;
determining that the second object is a whitelisted item;
based on the determination that the second object is a whitelisted item, reducing a percentage that the second object counts toward the threshold amount of the blacklisted item by a predetermined amount;
determining whether a combination of the first object and the second object represents more than a threshold amount of the blacklisted item;
denying the request to print the first object if the combination of the first object and the second object represents more than a threshold amount of the blacklisted item; and
allowing the request to print the first object if the combination of the first object and the second object does not represent more than a threshold amount of the blacklisted item.
16. The system of claim 15 , wherein applying the second model to the first object includes using a data table with a print history of a second profile linked to the first user profile based on shared communications between the first user profile and the second user profile.
17. The system of claim 15 , wherein:
applying the first model includes applying a tag to the first object based on a determined classification, the tag corresponding to the blacklisted item, and
applying the second model includes determining that the second object also includes the tag and determining an amount of the blacklisted item that the combination of the first object and the second object represents.
18. The system of claim 15 , wherein a print history of the first user profile is stored in the data table, wherein the print history includes a history of objects printed using the first user profile, and the stages further comprise comparing the data table to a blockchain record of user print histories to verify that no records have been removed from the table.
19. The system of claim 15 , the stages further comprising sending an alert to an administrator that identifies the combination of objects used to deny the request.
20. The system of claim 15 , wherein:
the first model and the second model are trained using machine learning; and
the stages further comprise:
retraining at least the first model or the second model based on a determination whether a combination of the first object and the second object represents more than a threshold amount of the blacklisted item.