Re: Algorithms for TCAS

From: (James K. Kuchar)
Organization: Massachusetts Institute of Technology
Date:         23 Feb 94 12:05:39 PST
References:   1
View raw article
  or MIME structure

In article <airliners.1994.941@orchard.Chicago.COM>, writes:
|>   I understand that the collision detection and avoidance 
|> algorithms used in TCAS products were developed by MITRE
|> Corporation for the FAA. I also understand some TCAS systems, when
|> operated around airports, tend to be a little paranoid and warn of
|> possible collisions when in fact there is no threat.  Better to be
|> safe than sorry I suppose.  Does anyone know if the TCAS
|> algorithms employ any mechanism for reasoning with uncertainty,
|> e.g., fuzzy logic, belief networks?  Also, can someone please
|> direct me to the correct group within MITRE to ask this question
|> directly?

Current TCAS specs are found in RTCA DO-185 (that's Radio Technical Commission
for Aeronautics), "Minimum Operational Performance Standards for TCAS".
DO-185 has the algorithms etc. for TCAS II which is what is
currently on civil jet transports. TCAS II uses range, range rate, altitude, and
altitude rate information to determine if an alert should be issued. All TCAS II
alerts give vertical guidance (climb, descend, etc.).

TCAS III was proposed a while back & would use lateral information also (bearing 
& bearing rate). A recent report by Lincoln Labs, however, has shown that TCAS 
III is unfeasible given current plans for sensor data accuracy. Effort is now
being focused on TCAS IV, which will use datalink of aircraft flight management
computer data (latitude, longitude, altitude, I think) to increase the accuracy of
position estimates.

I've talked to several people at Lincoln and Mitre who you may want to contact
as well for more info on TCAS III or IV (I don't know how much they've worked on

At Lincoln:
Vince Orlando (617)-981-7428
Doug Burgess  (617)-981-3610
Loren Wood             -3380

At Mitre:
Ned Spencer   (703)-883-6463

Hope this helps some.

Jim Kuchar