-
Notifications
You must be signed in to change notification settings - Fork 4
Home
MakeIt Labs has experienced a 100%+ growth in membership in a span of about 2 years. In the past, word-of-mouth and the honor system were just barely sufficient to control access to much of the equipment and tools available at the Lab. A few pieces of equipment had RFID lockouts, but most did not. These systems were either standalone systems (e.g. laser cutter, MakerBot), or integrated with the now-defunct legacy RFID system (short-lived wood shop door control).
As the membership has grown, the organization has adapted by instituting Resource Managers. These individuals are responsible for managing areas or individual pieces of equipment. Resource Managers are responsible for setting policies, training members, and maintaining the equipment for their area. With the drastic increase in membership size, incidents of equipment misuse and unreported damage have increased as the equipment sees more overall usage. Centrally managed, fine-grained electronic lockout/control of equipment has been a long-term desire that has finally risen to the top of the priority queue.
Coincident with the move to the new building in early 2016, the legacy RFID control systems (based on Google Apps API) that controlled building access were phased out and replaced. The old system had numerous problems not worth discussing here, but of note is that the new system was designed with the eventual goal of incorporating equipment/tool access with a fine level of control over resources and users. The system is based around a database-driven backend (see authbackend) that resource access control nodes can periodically poll to obtain Access Control Lists (ACL). The first access control nodes developed were based on Raspberry Pi, and have been reliably controlling building access since early 2016 (see doorbot).
The overarching goal of RATT is to create a low-cost access control node that builds on the work that has already been done on the new backend. The nodes will be a proven and tested platform of hardware and firmware providing enough flexibility to make them adaptable to a variety of applications at the Lab. Each tool or piece of equipment has unique needs that require custom logic or hardware to create a safe, robust and reliable access control method.
The RATT platform standardizes the elements that do not change across applications (e.g. wifi networking, ACL download via encrypted/authenticated https, event logging, robustness to network outages via non-volatile ACL/log caching, user interface elements such as display & buttons, etc.). A corollary output of the RATT project is to analyze various access control scenarios and create an application taxonomy. The aim is to categorize access control scenarios into a "reasonable" number of standardized models, and build reusable hardware & firmware solutions that can be deployed against these scenarios when additional tools/equipment are brought under access control.
While the existing Raspberry Pi 2/Pi 3 platform could be used and built upon for this purpose, it has a much higher cost per access control unit than other potential solutions, at $35-$40 per node. General interest in the inexpensive ($3/node) ESP8266 wifi microcontroller platform sparked interest in developing a system based around similar technology. In the intervening time, the much-improved ESP32 platform has been launched at slightly higher price point ($8-10/node), but still inexpensive and powerful enough for it to be the top contender for the project platform in early 2017.
However, as of January 2018, more options are now available in Pi-like platforms, including the Raspberry Pi Zero W ($10-$20/node street price, availability may be challenging due to supply) and the Orange Pi Zero ($14-$20/node street price). This has tilted the scale back in favor of a Pi-like solution as it provides a much friendlier and time-efficient software development platform that is more robust an has more support both in terms of community as well as in-house knowledge. This is a critical design consideration for the RATT project, as it must be sustainable long-term, preferably with a lower barrier to understanding than a 'hardcore' embedded platform.